From: Chet Ramey Date: Mon, 3 Nov 2014 19:31:34 +0000 (-0500) Subject: commit bash-20141024 snapshot X-Git-Tag: bash-4.4-alpha~45 X-Git-Url: http://git.ipfire.org/?a=commitdiff_plain;h=e0722a04ecfd67657486678e4481d7d089dfbcec;p=thirdparty%2Fbash.git commit bash-20141024 snapshot --- diff --git a/CWRU/CWRU.chlog b/CWRU/CWRU.chlog index 9c2941de8..13f4ee594 100644 --- a/CWRU/CWRU.chlog +++ b/CWRU/CWRU.chlog @@ -7020,3 +7020,20 @@ jobs.h trap.c - initialize_traps: add call to get_original_tty_job_signals + + 10/22 + ----- +subst.c + - expand_string_for_rhs: when expanding in this context (rhs of a word + expansion or pattern removal), we don't perform word splitting, so + we don't want to split $* if IFS is empty. Fixes bug reported by + Stephane Chazelas + + 10/23 + ----- +subst.c + - param_expand: when expanding $* in a pattern context where the + expansion is quoted (Q_PATQUOTE), don't quote the expansion -- + the outer quotes don't make the characters in the expansion of + $* special. Posix interp 221. Reported by Stephane Chazelas + diff --git a/CWRU/POSIX.NOTES.old b/CWRU/POSIX.NOTES.old new file mode 100644 index 000000000..1707ab10c --- /dev/null +++ b/CWRU/POSIX.NOTES.old @@ -0,0 +1,82 @@ +Starting bash with the `--posix' command-line option or executing +`set -o posix' while bash is running will cause bash to conform more +closely to the Posix.2 standard by changing the behavior to match that +specified by Posix.2 in areas where the bash default differs. + +The following list is what's changed when `posix mode' is in effect: + +1. When a command in the hash table no longer exists, bash will re-search + $PATH to find the new location. This is also available with + `shopt -s checkhash'. + +2. The >& redirection does not redirect stdout and stderr. + +3. The message printed by the job control code and builtins when a job + exits with a non-zero status is `Done(status)'. + +4. Reserved words may not be aliased. + +5. The Posix.2 PS1 and PS2 expansions of `!' -> history number and + `!!' -> `!' are enabled, and parameter expansion is performed on + the value regardless of the setting of the `promptvars' option. + +6. Interactive comments are enabled by default. (Note that bash has + them on by default anyway.) + +7. The Posix.2 startup files are executed ($ENV) rather than the normal + bash files. + +8. Tilde expansion is only performed on assignments preceding a command + name, rather than on all assignment statements on the line. + +9. The default history file is ~/.sh_history (default value of $HISTFILE). + +10. The output of `kill -l' prints all the signal names on a single line, + separated by spaces. + +11. Non-interactive shells exit if `file' in `. file' is not found. + +12. Redirection operators do not perform pathname expansion on the word + in the redirection unless the shell is interactive + +13. Function names must be valid shell identifiers. That is, they may not + contain characters other than letters, digits, and underscores, and + may not start with a digit. Declaring a function with an illegal name + causes a fatal syntax error in non-interactive shells. + +14. Posix.2 `special' builtins are found before shell functions during command + lookup. + +15. If a Posix.2 special builtin returns an error status, a non-interactive + shell exits. The fatal errors are those listed in the POSIX.2 standard, + and include things like passing incorrect options, redirection errors, + variable assignment errors for assignments preceding the command name, + and so on. + +16. The environment passed to executed commands is not sorted. Neither is + the output of `set'. This is not strictly Posix.2 behavior, but sh + does it this way. Ksh does not. It's not necessary to sort the + environment; no program should rely on it being sorted. + +17. If the `cd' builtin finds a directory to change to using $CDPATH, the + value it assigns to $PWD does not contain any symbolic links, as if + `cd -P' had been executed. + +18. A non-interactive shell exits with an error status if a variable + assignment error occurs when no command name follows the assignment + statements. A variable assignment error occurs, for example, when + trying to assign a value to a read-only variable. + +19. A non-interactive shell exits with an error status if the iteration + variable in a for statement or the selection variable in a select + statement is a read-only variable. + +20. Process substitution is not available. + +21. Assignment statements preceding POSIX.2 `special' builtins persist in + the shell environment after the builtin completes. + +There is other Posix.2 behavior that bash does not implement. Specifically: + +1. Assignment statements affect the execution environment of all builtins, + not just special ones. diff --git a/CWRU/old/set.def.save b/CWRU/old/set.def.save new file mode 100644 index 000000000..87b78d7cc --- /dev/null +++ b/CWRU/old/set.def.save @@ -0,0 +1,544 @@ +This file is set.def, from which is created set.c. +It implements the "set" and "unset" builtins in Bash. + +Copyright (C) 1987, 1989, 1991 Free Software Foundation, Inc. + +This file is part of GNU Bash, the Bourne Again SHell. + +Bash is free software; you can redistribute it and/or modify it under +the terms of the GNU General Public License as published by the Free +Software Foundation; either version 1, or (at your option) any later +version. + +Bash is distributed in the hope that it will be useful, but WITHOUT ANY +WARRANTY; without even the implied warranty of MERCHANTABILITY or +FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License +for more details. + +You should have received a copy of the GNU General Public License along +with Bash; see the file COPYING. If not, write to the Free Software +Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. + +$PRODUCES set.c + +#include +#include "../shell.h" +#include "../flags.h" + +#include "bashgetopt.h" + +extern int interactive; +extern int noclobber, posixly_correct; +#if defined (READLINE) +extern int rl_editing_mode, no_line_editing; +#endif /* READLINE */ + +$BUILTIN set +$FUNCTION set_builtin +$SHORT_DOC set [--abefhkmnptuvxldBCHP] [-o option] [arg ...] + -a Mark variables which are modified or created for export. + -b Notify of job termination immediately. + -e Exit immediately if a command exits with a non-zero status. + -f Disable file name generation (globbing). + -h Locate and remember function commands as functions are + defined. Function commands are normally looked up when + the function is executed. + -i Force the shell to be an "interactive" one. Interactive shells + always read `~/.bashrc' on startup. + -k All keyword arguments are placed in the environment for a + command, not just those that precede the command name. + -m Job control is enabled. + -n Read commands but do not execute them. + -o option-name + Set the variable corresponding to option-name: + allexport same as -a + braceexpand same as -B +#if defined (READLINE) + emacs use an emacs-style line editing interface +#endif /* READLINE */ + errexit same as -e + histexpand same as -H + ignoreeof the shell will not exit upon reading EOF + interactive-comments + allow comments to appear in interactive commands + monitor same as -m + noclobber disallow redirection to existing files + noexec same as -n + noglob same as -f + nohash same as -d + notify save as -b + nounset same as -u + physical same as -P + posix change the behavior of bash where the default + operation differs from the 1003.2 standard to + match the standard + privileged same as -p + verbose same as -v +#if defined (READLINE) + vi use a vi-style line editing interface +#endif /* READLINE */ + xtrace same as -x + -p Turned on whenever the real and effective user ids do not match. + Disables processing of the $ENV file and importing of shell + functions. Turning this option off causes the effective uid and + gid to be set to the real uid and gid. + -t Exit after reading and executing one command. + -u Treat unset variables as an error when substituting. + -v Print shell input lines as they are read. + -x Print commands and their arguments as they are executed. + -l Save and restore the binding of the NAME in a FOR command. + -d Disable the hashing of commands that are looked up for execution. + Normally, commands are remembered in a hash table, and once + found, do not have to be looked up again. +#if defined (BRACE_EXPANSION) + -B the shell will perform brace expansion +#endif /* BRACE_EXPANSION */ +#if defined (BANG_HISTORY) + -H Enable ! style history substitution. This flag is on + by default. +#endif /* BANG_HISTORY */ + -C If set, disallow existing regular files to be overwritten + by redirection of output. + -P If set, do not follow symbolic links when executing commands + such as cd which change the current directory. + +Using + rather than - causes these flags to be turned off. The +flags can also be used upon invocation of the shell. The current +set of flags may be found in $-. The remaining n ARGs are positional +parameters and are assigned, in order, to $1, $2, .. $n. If no +ARGs are given, all shell variables are printed. +$END + +/* An a-list used to match long options for set -o to the corresponding + option letter. */ +struct { + char *name; + int letter; +} o_options[] = { + { "allexport", 'a' }, +#if defined (BRACE_EXPANSION) + { "braceexpand",'B' }, +#endif + { "errexit", 'e' }, + { "histexpand", 'H' }, + { "monitor", 'm' }, + { "noexec", 'n' }, + { "noglob", 'f' }, + { "nohash", 'd' }, +#if defined (JOB_CONTROL) + { "notify", 'b' }, +#endif /* JOB_CONTROL */ + {"nounset", 'u' }, + {"physical", 'P' }, + {"privileged", 'p' }, + {"verbose", 'v' }, + {"xtrace", 'x' }, + {(char *)NULL, 0}, +}; + +#define MINUS_O_FORMAT "%-15s\t%s\n" + +void +list_minus_o_opts () +{ + register int i; + char *on = "on", *off = "off"; + + printf (MINUS_O_FORMAT, "noclobber", (noclobber == 1) ? on : off); + + if (find_variable ("ignoreeof") || find_variable ("IGNOREEOF")) + printf (MINUS_O_FORMAT, "ignoreeof", on); + else + printf (MINUS_O_FORMAT, "ignoreeof", off); + + printf (MINUS_O_FORMAT, "interactive-comments", + interactive_comments ? on : off); + + printf (MINUS_O_FORMAT, "posix", posixly_correct ? on : off); + +#if defined (READLINE) + if (no_line_editing) + { + printf (MINUS_O_FORMAT, "emacs", off); + printf (MINUS_O_FORMAT, "vi", off); + } + else + { + /* Magic. This code `knows' how readline handles rl_editing_mode. */ + printf (MINUS_O_FORMAT, "emacs", (rl_editing_mode == 1) ? on : off); + printf (MINUS_O_FORMAT, "vi", (rl_editing_mode == 0) ? on : off); + } +#endif /* READLINE */ + + for (i = 0; o_options[i].name; i++) + { + int *on_or_off, zero = 0; + + on_or_off = find_flag (o_options[i].letter); + if (on_or_off == FLAG_UNKNOWN) + on_or_off = &zero; + printf (MINUS_O_FORMAT, o_options[i].name, (*on_or_off == 1) ? on : off); + } +} + +set_minus_o_option (on_or_off, option_name) + int on_or_off; + char *option_name; +{ + int option_char = -1; + + if (STREQ (option_name, "noclobber")) + { + if (on_or_off == FLAG_ON) + bind_variable ("noclobber", ""); + else + unbind_variable ("noclobber"); + stupidly_hack_special_variables ("noclobber"); + } + else if (STREQ (option_name, "ignoreeof")) + { + unbind_variable ("ignoreeof"); + unbind_variable ("IGNOREEOF"); + if (on_or_off == FLAG_ON) + bind_variable ("IGNOREEOF", "10"); + stupidly_hack_special_variables ("IGNOREEOF"); + } + +#if defined (READLINE) + else if ((STREQ (option_name, "emacs")) || (STREQ (option_name, "vi"))) + { + if (on_or_off == FLAG_ON) + { + rl_variable_bind ("editing-mode", option_name); + + if (interactive) + with_input_from_stdin (); + no_line_editing = 0; + } + else + { + int isemacs = (rl_editing_mode == 1); + if ((isemacs && STREQ (option_name, "emacs")) || + (!isemacs && STREQ (option_name, "vi"))) + { + if (interactive) + with_input_from_stream (stdin, "stdin"); + no_line_editing = 1; + } + else + builtin_error ("not in %s editing mode", option_name); + } + } +#endif /* READLINE */ + else if (STREQ (option_name, "interactive-comments")) + interactive_comments = (on_or_off == FLAG_ON); + else if (STREQ (option_name, "posix")) + { + posixly_correct = (on_or_off == FLAG_ON); + unbind_variable ("POSIXLY_CORRECT"); + unbind_variable ("POSIX_PEDANTIC"); + if (on_or_off == FLAG_ON) + { + bind_variable ("POSIXLY_CORRECT", ""); + stupidly_hack_special_variables ("POSIXLY_CORRECT"); + } + } + else + { + register int i; + for (i = 0; o_options[i].name; i++) + { + if (STREQ (option_name, o_options[i].name)) + { + option_char = o_options[i].letter; + break; + } + } + if (option_char == -1) + { + builtin_error ("%s: unknown option name", option_name); + return (EXECUTION_FAILURE); + } + if (change_flag (option_char, on_or_off) == FLAG_ERROR) + { + bad_option (option_name); + return (EXECUTION_FAILURE); + } + } + return (EXECUTION_SUCCESS); +} + +/* Set some flags from the word values in the input list. If LIST is empty, + then print out the values of the variables instead. If LIST contains + non-flags, then set $1 - $9 to the successive words of LIST. */ +set_builtin (list) + WORD_LIST *list; +{ + int on_or_off, flag_name, force_assignment = 0; + + if (!list) + { + SHELL_VAR **vars; + + vars = all_shell_variables (); + if (vars) + { + print_var_list (vars); + free (vars); + } + + vars = all_shell_functions (); + if (vars) + { + print_var_list (vars); + free (vars); + } + + return (EXECUTION_SUCCESS); + } + + /* Check validity of flag arguments. */ + if (*list->word->word == '-' || *list->word->word == '+') + { + register char *arg; + WORD_LIST *save_list = list; + + while (list && (arg = list->word->word)) + { + char c; + + if (arg[0] != '-' && arg[0] != '+') + break; + + /* `-' or `--' signifies end of flag arguments. */ + if (arg[0] == '-' && + (!arg[1] || (arg[1] == '-' && !arg[2]))) + break; + + while (c = *++arg) + { + if (find_flag (c) == FLAG_UNKNOWN && c != 'o') + { + char s[2]; + s[0] = c; s[1] = '\0'; + bad_option (s); + if (c == '?') + builtin_usage (); + return (c == '?' ? EXECUTION_SUCCESS : EXECUTION_FAILURE); + } + } + list = list->next; + } + list = save_list; + } + + /* Do the set command. While the list consists of words starting with + '-' or '+' treat them as flags, otherwise, start assigning them to + $1 ... $n. */ + while (list) + { + char *string = list->word->word; + + /* If the argument is `--' or `-' then signal the end of the list + and remember the remaining arguments. */ + if (string[0] == '-' && (!string[1] || (string[1] == '-' && !string[2]))) + { + list = list->next; + + /* `set --' unsets the positional parameters. */ + if (string[1] == '-') + force_assignment = 1; + + /* Until told differently, the old shell behaviour of + `set - [arg ...]' being equivalent to `set +xv [arg ...]' + stands. Posix.2 says the behaviour is marked as obsolescent. */ + else + { + change_flag ('x', '+'); + change_flag ('v', '+'); + } + + break; + } + + if ((on_or_off = *string) && + (on_or_off == '-' || on_or_off == '+')) + { + int i = 1; + while (flag_name = string[i++]) + { + if (flag_name == '?') + { + builtin_usage (); + return (EXECUTION_SUCCESS); + } + else if (flag_name == 'o') /* -+o option-name */ + { + char *option_name; + WORD_LIST *opt; + + opt = list->next; + + if (!opt) + { + list_minus_o_opts (); + continue; + } + + option_name = opt->word->word; + + if (!option_name || !*option_name || (*option_name == '-')) + { + list_minus_o_opts (); + continue; + } + list = list->next; /* Skip over option name. */ + + if (set_minus_o_option (on_or_off, option_name) != EXECUTION_SUCCESS) + return (EXECUTION_FAILURE); + } + else + { + if (change_flag (flag_name, on_or_off) == FLAG_ERROR) + { + char opt[3]; + opt[0] = on_or_off; + opt[1] = flag_name; + opt[2] = '\0'; + bad_option (opt); + builtin_usage (); + return (EXECUTION_FAILURE); + } + } + } + } + else + { + break; + } + list = list->next; + } + + /* Assigning $1 ... $n */ + if (list || force_assignment) + remember_args (list, 1); + return (EXECUTION_SUCCESS); +} + +$BUILTIN unset +$FUNCTION unset_builtin +$SHORT_DOC unset [-f] [-v] [name ...] +For each NAME, remove the corresponding variable or function. Given +the `-v', unset will only act on variables. Given the `-f' flag, +unset will only act on functions. With neither flag, unset first +tries to unset a variable, and if that fails, then tries to unset a +function. Some variables (such as PATH and IFS) cannot be unset; also +see readonly. +$END + +#define NEXT_VARIABLE() any_failed++; list = list->next; continue; + +unset_builtin (list) + WORD_LIST *list; +{ + int unset_function, unset_variable, unset_array, opt, any_failed; + char *name; + + unset_function = unset_variable = unset_array = any_failed = 0; + + reset_internal_getopt (); + while ((opt = internal_getopt (list, "fv")) != -1) + { + switch (opt) + { + case 'f': + unset_function = 1; + break; + case 'v': + unset_variable = 1; + break; + default: + builtin_usage (); + return (EXECUTION_FAILURE); + } + } + + list = loptend; + + if (unset_function && unset_variable) + { + builtin_error ("cannot simultaneously unset a function and a variable"); + return (EXECUTION_FAILURE); + } + + while (list) + { + SHELL_VAR *var; + int tem; +#if defined (ARRAY_VARS) + char *t; +#endif + + name = list->word->word; + +#if defined (ARRAY_VARS) + if (!unset_function && valid_array_reference (name)) + { + t = strchr (name, '['); + *t++ = '\0'; + unset_array++; + } +#endif + + var = unset_function ? find_function (name) : find_variable (name); + + if (var && !unset_function && non_unsettable_p (var)) + { + builtin_error ("%s: cannot unset", name); + NEXT_VARIABLE (); + } + + /* Posix.2 says that unsetting readonly variables is an error. */ + if (var && readonly_p (var)) + { + builtin_error ("%s: cannot unset: readonly %s", + name, unset_function ? "function" : "variable"); + NEXT_VARIABLE (); + } + + /* Unless the -f option is supplied, the name refers to a variable. */ +#if defined (ARRAY_VARS) + if (var && unset_array) + { + if (array_p (var) == 0) + { + builtin_error ("%s: not an array variable", name); + NEXT_VARIABLE (); + } + else + tem = unbind_array_element (var, t); + } + else +#endif /* ARRAY_VARS */ + tem = makunbound (name, unset_function ? shell_functions : shell_variables); + + /* This is what Posix.2 draft 11+ says. ``If neither -f nor -v + is specified, the name refers to a variable; if a variable by + that name does not exist, a function by that name, if any, + shall be unset.'' */ + if ((tem == -1) && !unset_function && !unset_variable) + tem = makunbound (name, shell_functions); + + if (tem == -1) + any_failed++; + else if (!unset_function) + stupidly_hack_special_variables (name); + + list = list->next; + } + + if (any_failed) + return (EXECUTION_FAILURE); + else + return (EXECUTION_SUCCESS); +} diff --git a/CWRU/save/unwind_prot.h.save b/CWRU/save/unwind_prot.h.save new file mode 100644 index 000000000..998fd72b6 --- /dev/null +++ b/CWRU/save/unwind_prot.h.save @@ -0,0 +1,50 @@ +/* unwind_prot.h - Macros and functions for hacking unwind protection. */ + +/* Copyright (C) 1993 Free Software Foundation, Inc. + + This file is part of GNU Bash, the Bourne Again SHell. + + Bash is free software; you can redistribute it and/or modify it under + the terms of the GNU General Public License as published by the Free + Software Foundation; either version 2, or (at your option) any later + version. + + Bash is distributed in the hope that it will be useful, but WITHOUT ANY + WARRANTY; without even the implied warranty of MERCHANTABILITY or + FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License + for more details. + + You should have received a copy of the GNU General Public License along + with Bash; see the file COPYING. If not, write to the Free Software + Foundation, 675 Mass Ave, Cambridge, MA 02139, USA. */ + +#if !defined (_UNWIND_PROT_H) +#define _UNWIND_PROT_H + +/* Run a function without interrupts. */ +extern void begin_unwind_frame (); +extern void discard_unwind_frame (); +extern void run_unwind_frame (); +extern void add_unwind_protect (); +extern void remove_unwind_protect (); +extern void run_unwind_protects (); +extern void unwind_protect_var (); + +/* Define for people who like their code to look a certain way. */ +#define end_unwind_frame() + +/* How to protect an integer. */ +#define unwind_protect_int(X) unwind_protect_var (&(X), (char *)(X), sizeof (int)) + +/* How to protect a pointer to a string. */ +#define unwind_protect_string(X) \ + unwind_protect_var ((int *)&(X), (X), sizeof (char *)) + +/* How to protect any old pointer. */ +#define unwind_protect_pointer(X) unwind_protect_string (X) + +/* How to protect the contents of a jmp_buf. */ +#define unwind_protect_jmp_buf(X) \ + unwind_protect_var ((int *)(X), (char *)(X), sizeof (procenv_t)) + +#endif /* _UNWIND_PROT_H */ diff --git a/MANIFEST b/MANIFEST index e67b9c737..fb21780e1 100644 --- a/MANIFEST +++ b/MANIFEST @@ -898,6 +898,7 @@ tests/dollar-at-star f tests/dollar-at-star1.sub f tests/dollar-at-star2.sub f tests/dollar-at-star3.sub f +tests/dollar-at-star4.sub f tests/dollar-at1.sub f tests/dollar-at2.sub f tests/dollar-at3.sub f diff --git a/MANIFEST~ b/MANIFEST~ new file mode 100644 index 000000000..fd03646ec --- /dev/null +++ b/MANIFEST~ @@ -0,0 +1,1304 @@ +# +# Master distribution manifest for bash +# +# +# Filename type +# +CWRU d +CWRU/misc d +builtins d +cross-build d +doc d +examples d +#examples/obashdb d +examples/complete d +examples/functions d +examples/scripts d +#examples/scripts.v2 d +#examples/scripts.noah d +examples/startup-files d +#examples/startup-files/apple d +examples/misc d +examples/loadables d +examples/loadables/perl d +include d +lib d +lib/glob d +lib/glob/doc d +lib/intl d +lib/malloc d +lib/readline d +lib/readline/doc d +lib/readline/examples d +lib/sh d +lib/termcap d +lib/tilde d +m4 d +po d +support d +tests d +tests/misc d +ABOUT-NLS f +ChangeLog s CWRU/changelog +CHANGES f +COMPAT f +COPYING f +INSTALL f +MANIFEST f +NEWS f +NOTES f +POSIX f +README f +RBASH f +AUTHORS f +Y2K f +configure.ac f +configure f 755 +Makefile.in f +config-top.h f +config-bot.h f +config.h.in f +aclocal.m4 f +array.c f +arrayfunc.c f +assoc.c f +eval.c f +print_cmd.c f +general.c f +list.c f +locale.c f +stringlib.c f +variables.c f +make_cmd.c f +copy_cmd.c f +unwind_prot.c f +dispose_cmd.c f +bashhist.c f +hashcmd.c f +hashlib.c f +parse.y f +pathexp.c f +subst.c f +shell.c f +trap.c f +sig.c f +siglist.c f +version.c f +flags.c f +jobs.c f +input.c f +mailcheck.c f +test.c f +expr.c f +alias.c f +execute_cmd.c f +findcmd.c f +redir.c f +bashline.c f +braces.c f +bracecomp.c f +nojobs.c f +error.c f +xmalloc.c f +pcomplete.c f +pcomplib.c f +mksyntax.c f +alias.h f +builtins.h f +bashhist.h f +bashline.h f +conftypes.h f +patchlevel.h f +variables.h f +array.h f +arrayfunc.h f +assoc.h f +jobs.h f +findcmd.h f +hashlib.h f +quit.h f +flags.h f +shell.h f +syntax.h f +pathexp.h f +parser.h f +pcomplete.h f +sig.h f +test.h f +trap.h f +general.h f +unwind_prot.h f +input.h f +error.h f +command.h f +externs.h f +siglist.h f +subst.h f +dispose_cmd.h f +hashcmd.h f +bashansi.h f +bashjmp.h f +bashintl.h f +make_cmd.h f +execute_cmd.h f +redir.h f +bashtypes.h f +mailcheck.h f +xmalloc.h f +y.tab.c f +y.tab.h f +parser-built f +pathnames.h.in f +builtins/Makefile.in f +builtins/alias.def f +builtins/bind.def f +builtins/break.def f +builtins/builtin.def f +builtins/caller.def f +builtins/cd.def f +builtins/colon.def f +builtins/command.def f +builtins/complete.def f +builtins/common.c f +builtins/declare.def f +builtins/echo.def f +builtins/enable.def f +builtins/eval.def f +builtins/evalfile.c f +builtins/evalstring.c f +builtins/exec.def f +builtins/exit.def f +builtins/fc.def f +builtins/fg_bg.def f +builtins/gen-helpfiles.c f +builtins/getopt.c f +builtins/getopt.h f +builtins/getopts.def f +builtins/hash.def f +builtins/help.def f +builtins/let.def f +builtins/history.def f +builtins/jobs.def f +builtins/kill.def f +builtins/mapfile.def f +builtins/mkbuiltins.c f +builtins/printf.def f +builtins/pushd.def f +builtins/read.def f +builtins/reserved.def f +builtins/return.def f +builtins/set.def f +builtins/setattr.def f +builtins/shift.def f +builtins/shopt.def f +builtins/source.def f +builtins/suspend.def f +builtins/test.def f +builtins/times.def f +builtins/trap.def f +builtins/type.def f +builtins/ulimit.def f +builtins/umask.def f +builtins/wait.def f +builtins/psize.c f +builtins/psize.sh f +builtins/inlib.def f +builtins/bashgetopt.c f +builtins/common.h f +builtins/bashgetopt.h f +cross-build/cygwin32.cache f +cross-build/x86-beos.cache f +cross-build/opennt.cache f +include/ansi_stdlib.h f +include/chartypes.h f +include/filecntl.h f +include/gettext.h f +include/maxpath.h f +include/memalloc.h f +include/ocache.h f +include/posixdir.h f +include/posixjmp.h f +include/posixselect.h f +include/posixstat.h f +include/posixtime.h f +include/posixwait.h f +include/shmbchar.h f +include/shmbutil.h f +include/shtty.h f +include/stat-time.h f +include/stdc.h f +include/systimes.h f +include/typemax.h f +include/unionwait.h f +lib/glob/Makefile.in f +lib/glob/sm_loop.c f +lib/glob/smatch.c f +lib/glob/strmatch.c f +lib/glob/strmatch.h f +lib/glob/glob.c f +lib/glob/glob.h f +lib/glob/glob_loop.c f +lib/glob/gmisc.c f +lib/glob/xmbsrtowcs.c f +lib/glob/collsyms.h f +lib/glob/doc/Makefile f +lib/glob/doc/glob.texi f +lib/glob/ndir.h f +lib/intl/ChangeLog f +lib/intl/Makefile.in f +lib/intl/VERSION f +lib/intl/bindtextdom.c f +lib/intl/config.charset f +lib/intl/dcgettext.c f +lib/intl/dcigettext.c f +lib/intl/dcngettext.c f +lib/intl/dgettext.c f +lib/intl/dngettext.c f +lib/intl/eval-plural.h f +lib/intl/explodename.c f +lib/intl/finddomain.c f +lib/intl/gettext.c f +lib/intl/gettextP.h f +lib/intl/gmo.h f +lib/intl/hash-string.h f +lib/intl/intl-compat.c f +lib/intl/l10nflist.c f +lib/intl/libgnuintl.h.in f +lib/intl/loadinfo.h f +lib/intl/loadmsgcat.c f +lib/intl/localcharset.c f +lib/intl/localcharset.h f +lib/intl/locale.alias f +lib/intl/localealias.c f +lib/intl/localename.c f +lib/intl/log.c f +lib/intl/ngettext.c f +lib/intl/os2compat.c f +lib/intl/os2compat.h f +lib/intl/osdep.c f +lib/intl/plural-exp.c f +lib/intl/plural-exp.h f +lib/intl/plural.c f +lib/intl/plural.y f +lib/intl/ref-add.sin f +lib/intl/ref-del.sin f +lib/intl/relocatable.c f +lib/intl/relocatable.h f +lib/intl/textdomain.c f +lib/malloc/Makefile.in f +lib/malloc/getpagesize.h f +lib/malloc/imalloc.h f +lib/malloc/mstats.h f +lib/malloc/shmalloc.h f +lib/malloc/table.h f +lib/malloc/watch.h f +lib/malloc/alloca.c f +lib/malloc/malloc.c f +lib/malloc/stats.c f +lib/malloc/table.c f +lib/malloc/trace.c f +lib/malloc/watch.c f +lib/malloc/xmalloc.c f +lib/malloc/xleaktrace f 755 +lib/malloc/stub.c f +lib/malloc/i386-alloca.s f +lib/malloc/x386-alloca.s f +lib/readline/COPYING f +lib/readline/Makefile.in f +lib/readline/ChangeLog f +lib/readline/README f +lib/readline/STANDALONE f +lib/readline/readline.c f +lib/readline/vi_mode.c f +lib/readline/emacs_keymap.c f +lib/readline/vi_keymap.c f +lib/readline/history.c f +lib/readline/histexpand.c f +lib/readline/histsearch.c f +lib/readline/histfile.c f +lib/readline/funmap.c f +lib/readline/keymaps.c f +lib/readline/util.c f +lib/readline/terminal.c f +lib/readline/xfree.c f +lib/readline/xmalloc.c f +lib/readline/search.c f +lib/readline/isearch.c f +lib/readline/parens.c f +lib/readline/rltty.c f +lib/readline/compat.c f +lib/readline/complete.c f +lib/readline/bind.c f +lib/readline/display.c f +lib/readline/signals.c f +lib/readline/kill.c f +lib/readline/text.c f +lib/readline/undo.c f +lib/readline/macro.c f +lib/readline/input.c f +lib/readline/callback.c f +lib/readline/mbutil.c f +lib/readline/misc.c f +lib/readline/nls.c f +lib/readline/shell.c f +lib/readline/colors.c f +lib/readline/parse-colors.c f +lib/readline/savestring.c f +lib/readline/tilde.c f +lib/readline/tilde.h f +lib/readline/rldefs.h f +lib/readline/rlconf.h f +lib/readline/rlmbutil.h f +lib/readline/rlshell.h f +lib/readline/rltty.h f +lib/readline/rltypedefs.h f +lib/readline/rlwinsize.h f +lib/readline/readline.h f +lib/readline/tcap.h f +lib/readline/keymaps.h f +lib/readline/history.h f +lib/readline/histlib.h f +lib/readline/chardefs.h f +lib/readline/posixdir.h f +lib/readline/posixjmp.h f +lib/readline/posixselect.h f +lib/readline/posixstat.h f +lib/readline/ansi_stdlib.h f +lib/readline/rlstdc.h f +lib/readline/rlprivate.h f +lib/readline/colors.h f +lib/readline/parse-colors.h f +lib/readline/xmalloc.h f +lib/readline/doc/Makefile f +lib/readline/doc/version.texi f +lib/readline/doc/rlman.texi f +lib/readline/doc/rltech.texi f +lib/readline/doc/rluser.texi f +lib/readline/doc/rluserman.texi f +lib/readline/doc/history.texi f +lib/readline/doc/hstech.texi f +lib/readline/doc/hsuser.texi f +lib/readline/doc/fdl.texi f +lib/readline/examples/Makefile f +lib/readline/examples/excallback.c f +lib/readline/examples/fileman.c f +lib/readline/examples/manexamp.c f +lib/readline/examples/histexamp.c f +lib/readline/examples/rltest.c f +lib/readline/examples/rl-callbacktest.c f +lib/readline/examples/rl.c f +lib/readline/examples/rlcat.c f +lib/readline/examples/Inputrc f +lib/sh/Makefile.in f +lib/sh/casemod.c f +lib/sh/clktck.c f +lib/sh/clock.c f +lib/sh/dprintf.c f +lib/sh/eaccess.c f +lib/sh/fmtullong.c f +lib/sh/fmtulong.c f +lib/sh/fmtumax.c f +lib/sh/fnxform.c f +lib/sh/fpurge.c f +lib/sh/getcwd.c f +lib/sh/getenv.c f +lib/sh/inet_aton.c f +lib/sh/input_avail.c f +lib/sh/itos.c f +lib/sh/mailstat.c f +lib/sh/makepath.c f +lib/sh/mbscasecmp.c f +lib/sh/mbschr.c f +lib/sh/mbscmp.c f +lib/sh/memset.c f +lib/sh/mktime.c f +lib/sh/netconn.c f +lib/sh/netopen.c f +lib/sh/oslib.c f +lib/sh/pathcanon.c f +lib/sh/pathphys.c f +lib/sh/rename.c f +lib/sh/setlinebuf.c f +lib/sh/shmatch.c f +lib/sh/shmbchar.c f +lib/sh/shquote.c f +lib/sh/shtty.c f +lib/sh/snprintf.c f +lib/sh/spell.c f +lib/sh/strcasecmp.c f +lib/sh/strcasestr.c f +lib/sh/strchrnul.c f +lib/sh/strdup.c f +lib/sh/strerror.c f +lib/sh/strftime.c f +lib/sh/stringlist.c f +lib/sh/stringvec.c f +lib/sh/strnlen.c f +lib/sh/strpbrk.c f +lib/sh/strstr.c f +lib/sh/strtod.c f +lib/sh/strtoimax.c f +lib/sh/strtol.c f +lib/sh/strtoll.c f +lib/sh/strtoul.c f +lib/sh/strtoull.c f +lib/sh/strtoumax.c f +lib/sh/strtrans.c f +lib/sh/times.c f +lib/sh/timeval.c f +lib/sh/tmpfile.c f +lib/sh/uconvert.c f +lib/sh/ufuncs.c f +lib/sh/unicode.c f +lib/sh/vprint.c f +lib/sh/wcsdup.c f +lib/sh/wcsnwidth.c f +lib/sh/wcswidth.c f +lib/sh/winsize.c f +lib/sh/zcatfd.c f +lib/sh/zgetline.c f +lib/sh/zmapfd.c f +lib/sh/zread.c f +lib/sh/zwrite.c f +lib/termcap/Makefile.in f +lib/termcap/ltcap.h f +lib/termcap/termcap.c f +lib/termcap/termcap.h f +lib/termcap/tparam.c f +lib/termcap/version.c f +lib/tilde/README f +lib/tilde/Makefile.in f +lib/tilde/tilde.c f +lib/tilde/tilde.h f +lib/tilde/shell.c f +m4/stat-time.m4 f +m4/timespec.m4 f +po/LINGUAS f +po/Makefile.in.in f +po/Makevars f +po/POTFILES.in f +po/README f +po/Rules-builtins f +po/Rules-quot f +po/bash.pot f +po/boldquot.sed f +po/en@boldquot.gmo f +po/en@boldquot.header f +po/en@boldquot.po f +po/en@quot.gmo f +po/en@quot.header f +po/en@quot.po f +po/af.gmo f +po/af.po f +po/bg.gmo f +po/bg.po f +po/ca.gmo f +po/ca.po f +po/cs.gmo f +po/cs.po f +po/da.gmo f +po/da.po f +po/de.gmo f +po/de.po f +po/el.gmo f +po/el.po f +po/eo.gmo f +po/eo.po f +po/es.gmo f +po/es.po f +po/et.gmo f +po/et.po f +po/fi.gmo f +po/fi.po f +po/fr.gmo f +po/fr.po f +po/ga.gmo f +po/ga.po f +po/gl.gmo f +po/gl.po f +po/hr.gmo f +po/hr.po f +po/hu.gmo f +po/hu.po f +po/id.gmo f +po/id.po f +po/it.gmo f +po/it.po f +po/ja.gmo f +po/ja.po f +po/lt.gmo f +po/lt.po f +po/nl.gmo f +po/nl.po f +po/pl.gmo f +po/pl.po f +po/pt_BR.gmo f +po/pt_BR.po f +po/ro.gmo f +po/ro.po f +po/ru.gmo f +po/ru.po f +po/sk.gmo f +po/sk.po f +po/sl.gmo f +po/sr.po f +po/sr.gmo f +po/sl.po f +po/sv.gmo f +po/sv.po f +po/tr.gmo f +po/tr.po f +po/uk.gmo f +po/uk.po f +po/vi.gmo f +po/vi.po f +po/zh_CN.gmo f +po/zh_CN.po f +po/zh_TW.gmo f +po/zh_TW.po f +po/insert-header.sin f +po/quot.sed f +po/remove-potcdate.sin f +CWRU/misc/open-files.c f +CWRU/misc/sigs.c f +CWRU/misc/sigstat.c f +CWRU/misc/bison f +CWRU/misc/errlist.c f +CWRU/misc/hpux10-dlfcn.h f +CWRU/PLATFORMS f +CWRU/README f +CWRU/changelog f +CWRU/sh-redir-hack f +doc/FAQ f +doc/Makefile.in f +doc/bash.1 f +doc/bashbug.1 f +doc/builtins.1 f +doc/rbash.1 f +doc/README f +doc/INTRO f +doc/texinfo.tex f +doc/bashref.texi f +doc/version.texi f +doc/bashref.info f +doc/article.ms f +doc/htmlpost.sh f 755 +doc/infopost.sh f 755 +doc/fdl.texi f +doc/fdl.txt f +# +doc/article.ps f +doc/rose94.ps f +doc/bash.ps f +doc/bashbug.ps f +doc/builtins.ps f +doc/rbash.ps f +doc/bashref.ps f +doc/bashref.dvi f +doc/bash.0 f +doc/bashbug.0 f +doc/builtins.0 f +doc/rbash.0 f +doc/article.txt f +doc/bash.html f +doc/bashref.html f +doc/article.pdf f +doc/bash.pdf f +doc/bashref.pdf f +doc/rose94.pdf f +doc/aosa-bash.pdf f +# +support/Makefile.in f +support/bashversion.c f +support/checkbashisms f 755 +support/config.guess f +support/config.rpath f 755 +support/config.sub f +support/printenv.sh f 755 +support/printenv.c f +support/bash.xbm f +support/missing f 755 +support/mkclone f 755 +support/mkconffiles f 755 +support/mkdirs f 755 +support/mkinstalldirs f 755 +support/mkversion.sh f 755 +support/mksignames.c f +support/signames.c f +support/bashbug.sh f +support/man2html.c f +support/recho.c f +support/zecho.c f +support/xcase.c f +support/SYMLINKS f +support/fixlinks f 755 +support/install.sh f 755 +support/texi2dvi f 755 +support/texi2html f 755 +support/xenix-link.sh f 755 +support/shobj-conf f 755 +support/rlvers.sh f 755 +examples/INDEX.txt f +examples/INDEX.html f +#examples/obashdb/PERMISSION f +#examples/obashdb/README f +#examples/obashdb/bashdb f +#examples/obashdb/bashdb.el f +examples/complete/bash_completion f +examples/complete/cdfunc f +examples/complete/complete-examples f +#examples/complete/complete.ianmac f +#examples/complete/complete2.ianmac f +#examples/complete/complete.freebsd f +#examples/complete/complete.gnu-longopt f +examples/complete/bashcc-1.0.1.tar.gz f +examples/loadables/README f +examples/loadables/template.c f +examples/loadables/Makefile.in f +examples/loadables/necho.c f +examples/loadables/hello.c f +examples/loadables/print.c f +examples/loadables/realpath.c f +examples/loadables/sleep.c f +examples/loadables/strftime.c f +examples/loadables/truefalse.c f +#examples/loadables/getconf.h f +#examples/loadables/getconf.c f +examples/loadables/finfo.c f +examples/loadables/cat.c f +#examples/loadables/cut.c f +examples/loadables/logname.c f +examples/loadables/basename.c f +examples/loadables/dirname.c f +examples/loadables/tty.c f +examples/loadables/pathchk.c f +examples/loadables/tee.c f +examples/loadables/rmdir.c f +examples/loadables/head.c f +examples/loadables/printenv.c f +examples/loadables/push.c f +examples/loadables/id.c f +examples/loadables/whoami.c f +examples/loadables/uname.c f +examples/loadables/sync.c f +examples/loadables/mkdir.c f +examples/loadables/ln.c f +examples/loadables/mypid.c f +examples/loadables/unlink.c f +examples/loadables/perl/Makefile.in f +examples/loadables/perl/README f +examples/loadables/perl/bperl.c f +examples/loadables/perl/iperl.c f +#examples/loadables/sprintf.c f +#examples/loadables/xtitle.c f +examples/functions/array-stuff f +examples/functions/array-to-string f +examples/functions/autoload f +examples/functions/autoload.v2 f +examples/functions/autoload.v3 f +examples/functions/basename f +#examples/functions/basename2 f +#examples/functions/coproc.bash f +#examples/functions/coshell.README f +#examples/functions/coshell.bash f +examples/functions/csh-compat f +#examples/functions/dirfuncs f +examples/functions/dirname f +examples/functions/dirstack f +#examples/functions/emptydir f +examples/functions/exitstat f +examples/functions/external f +examples/functions/fact f +examples/functions/fstty f +examples/functions/func f +#examples/functions/gethtml f +#examples/functions/getoptx.bash f +examples/functions/inetaddr f +examples/functions/inpath f +#examples/functions/isnum.bash f +examples/functions/isnum2 f +examples/functions/isvalidip f +#examples/functions/jdate.bash f +#examples/functions/jj.bash f +#examples/functions/keep f +examples/functions/ksh-cd f +examples/functions/ksh-compat-test f +examples/functions/kshenv f +examples/functions/login f +#examples/functions/lowercase f +#examples/functions/manpage f +#examples/functions/mhfold f +#examples/functions/newdirstack.bsh f +examples/functions/notify.bash f +#examples/functions/pathfuncs f +#examples/functions/recurse f +#examples/functions/repeat2 f +#examples/functions/repeat3 f +examples/functions/seq f +examples/functions/seq2 f +examples/functions/shcat f +examples/functions/shcat2 f +examples/functions/sort-pos-params f +#examples/functions/sqroot f +examples/functions/substr f +examples/functions/substr2 f +#examples/functions/term f +examples/functions/whatis f +examples/functions/whence f +examples/functions/which f +#examples/functions/xalias.bash f +#examples/functions/xfind.bash f +#examples/scripts/adventure.sh f +#examples/scripts/bash-hexdump.sh f +#examples/scripts/bcsh.sh f +examples/scripts/cat.sh f +examples/scripts/center f +#examples/scripts/dd-ex.sh f +#examples/scripts/fixfiles.bash f +#examples/scripts/hanoi.bash f +examples/scripts/inpath f +#examples/scripts/krand.bash f +#examples/scripts/line-input.bash f +#examples/scripts/nohup.bash f +#examples/scripts/precedence f +#examples/scripts/randomcard.bash f +#examples/scripts/scrollbar f +#examples/scripts/scrollbar2 f +#examples/scripts/self-repro f +#examples/scripts/showperm.bash f +examples/scripts/shprompt f +examples/scripts/spin.bash f +#examples/scripts/timeout f +#examples/scripts/timeout2 f +#examples/scripts/timeout3 f +#examples/scripts/vtree2 f +#examples/scripts/vtree3 f +#examples/scripts/vtree3a f +#examples/scripts/websrv.sh f +examples/scripts/xterm_title f +examples/scripts/zprintf f +examples/startup-files/README f +examples/startup-files/Bashrc.bfox f +examples/startup-files/Bash_aliases f +examples/startup-files/Bash_profile f +examples/startup-files/bash-profile f +examples/startup-files/bashrc f +#examples/startup-files/apple/README f +#examples/startup-files/apple/aliases f +#examples/startup-files/apple/bash.defaults f +#examples/startup-files/apple/environment f +#examples/startup-files/apple/login f +#examples/startup-files/apple/logout f +#examples/startup-files/apple/rc f +#examples/misc/suncmd.termcap f +examples/misc/aliasconv.sh f +examples/misc/aliasconv.bash f +examples/misc/cshtobash f +tests/README f +tests/COPYRIGHT f +tests/alias.tests f +tests/alias1.sub f +tests/alias.right f +tests/appendop.tests f +tests/appendop1.sub f +tests/appendop2.sub f +tests/appendop.right f +tests/arith-for.tests f +tests/arith-for.right f +tests/arith.tests f +tests/arith.right f +tests/arith1.sub f +tests/arith2.sub f +tests/arith3.sub f +tests/arith4.sub f +tests/arith5.sub f +tests/arith6.sub f +tests/array.tests f +tests/array.right f +tests/array1.sub f +tests/array2.sub f +tests/array3.sub f +tests/array4.sub f +tests/array5.sub f +tests/array6.sub f +tests/array7.sub f +tests/array8.sub f +tests/array9.sub f +tests/array10.sub f +tests/array11.sub f +tests/array12.sub f +tests/array13.sub f +tests/array14.sub f +tests/array15.sub f +tests/array16.sub f +tests/array17.sub f +tests/array-at-star f +tests/array2.right f +tests/assoc.tests f +tests/assoc.right f +tests/assoc1.sub f +tests/assoc2.sub f +tests/assoc3.sub f +tests/assoc4.sub f +tests/assoc5.sub f +tests/assoc6.sub f +tests/assoc7.sub f +tests/braces.tests f +tests/braces.right f +tests/builtins.tests f +tests/builtins.right f +tests/builtins1.sub f +tests/builtins2.sub f +tests/builtins3.sub f +tests/builtins4.sub f +tests/source1.sub f +tests/source2.sub f +tests/source3.sub f +tests/source4.sub f +tests/source5.sub f +tests/source6.sub f +tests/source7.sub f +tests/case.tests f +tests/case.right f +tests/case1.sub f +tests/casemod.tests f +tests/casemod.right f +tests/comsub.tests f +tests/comsub.right f +tests/comsub1.sub f +tests/comsub-eof.tests f +tests/comsub-eof0.sub f +tests/comsub-eof1.sub f +tests/comsub-eof2.sub f +tests/comsub-eof3.sub f +tests/comsub-eof4.sub f +tests/comsub-eof5.sub f +tests/comsub-eof.right f +tests/comsub-posix.tests f +tests/comsub-posix.right f +tests/comsub-posix1.sub f +tests/comsub-posix2.sub f +tests/comsub-posix3.sub f +tests/cond.tests f +tests/cond.right f +tests/cond-regexp1.sub f +tests/cond-regexp2.sub f +tests/cond-regexp3.sub f +tests/coproc.tests f +tests/coproc.right f +tests/cprint.tests f +tests/cprint.right f +tests/dbg-support.right f +tests/dbg-support.sub f +tests/dbg-support.tests f +tests/dbg-support2.right f +tests/dbg-support2.tests f +tests/dbg-support3.sub f +tests/dollar-at-star f +tests/dollar-at-star1.sub f +tests/dollar-at-star2.sub f +tests/dollar-at-star3.sub f +tests/dollar-at1.sub f +tests/dollar-at2.sub f +tests/dollar-at3.sub f +tests/dollar-at4.sub f +tests/dollar-at5.sub f +tests/dollar-at6.sub f +tests/dollar-star1.sub f +tests/dollar-star2.sub f +tests/dollar-star3.sub f +tests/dollar-star4.sub f +tests/dollar-star5.sub f +tests/dollar-star6.sub f +tests/dollar-star7.sub f +tests/dollar.right f +tests/dstack.tests f +tests/dstack.right f +tests/dstack2.tests f +tests/dstack2.right f +tests/errors.tests f +tests/errors.right f +tests/errors1.sub f +tests/errors2.sub f +tests/errors3.sub f +tests/execscript f +tests/exec.right f +tests/exec1.sub f 755 +tests/exec2.sub f +tests/exec3.sub f +tests/exec4.sub f +tests/exec5.sub f +tests/exec6.sub f +tests/exec7.sub f +tests/exec8.sub f +tests/exec9.sub f +tests/exp.tests f +tests/exp.right f +tests/exp1.sub f +tests/exp2.sub f +tests/exp3.sub f +tests/exp4.sub f +tests/exp5.sub f +tests/exp6.sub f +tests/exportfunc.tests f +tests/exportfunc.right f +tests/exportfunc1.sub f +tests/exportfunc2.sub f +tests/exportfunc3.sub f +tests/extglob.tests f +tests/extglob.right f +tests/extglob1.sub f +tests/extglob1a.sub f +tests/extglob3.sub f +tests/extglob2.tests f +tests/extglob2.right f +tests/extglob3.tests f +tests/extglob3.right f +tests/func.tests f +tests/func.right f +tests/func1.sub f +tests/func2.sub f +tests/func3.sub f +tests/func4.sub f +tests/getopts.tests f +tests/getopts.right f +tests/getopts1.sub f +tests/getopts2.sub f +tests/getopts3.sub f +tests/getopts4.sub f +tests/getopts5.sub f +tests/getopts6.sub f +tests/getopts7.sub f +tests/glob.tests f +tests/glob1.sub f +tests/glob.right f +tests/globstar.tests f +tests/globstar.right f +tests/globstar1.sub f +tests/globstar2.sub f +tests/heredoc.tests f +tests/heredoc.right f +tests/heredoc1.sub f +tests/heredoc2.sub f +tests/heredoc3.sub f +tests/herestr.tests f +tests/herestr.right f +tests/histexp.tests f +tests/histexp.right f +tests/history.tests f +tests/history.right f +tests/history.list f 444 +tests/history1.sub f +tests/history2.sub f +tests/ifs.tests f +tests/ifs.right f +tests/ifs-posix.tests f +tests/ifs-posix.right f +tests/input-line.sh f +tests/input-line.sub f +tests/input.right f +tests/intl.tests f +tests/intl1.sub f +tests/intl2.sub f +tests/intl.right f +tests/iquote.tests f +tests/iquote.right f +tests/iquote1.sub f +tests/invert.tests f +tests/invert.right f +tests/jobs.tests f +tests/jobs1.sub f +tests/jobs2.sub f +tests/jobs3.sub f +tests/jobs4.sub f +tests/jobs5.sub f +tests/jobs.right f +tests/lastpipe.right f +tests/lastpipe.tests f +tests/lastpipe1.sub f +tests/mapfile.data f +tests/mapfile.right f +tests/mapfile.tests f +tests/mapfile1.sub f +tests/mapfile2.sub f +tests/more-exp.tests f +tests/more-exp.right f +tests/nameref.tests f +tests/nameref1.sub f +tests/nameref2.sub f +tests/nameref3.sub f +tests/nameref4.sub f +tests/nameref5.sub f +tests/nameref6.sub f +tests/nameref7.sub f +tests/nameref8.sub f +tests/nameref.right f +tests/new-exp.tests f +tests/new-exp1.sub f +tests/new-exp2.sub f +tests/new-exp3.sub f +tests/new-exp4.sub f +tests/new-exp5.sub f +tests/new-exp6.sub f +tests/new-exp7.sub f +tests/new-exp8.sub f +tests/new-exp9.sub f +tests/new-exp.right f +tests/nquote.tests f +tests/nquote.right f +tests/nquote1.sub f +tests/nquote2.sub f +tests/nquote1.tests f +tests/nquote1.right f +tests/nquote2.tests f +tests/nquote2.right f +tests/nquote3.tests f +tests/nquote3.right f +tests/nquote4.tests f +tests/nquote4.right f +tests/nquote5.tests f +tests/nquote5.right f +tests/parser.tests f +tests/parser.right f +tests/parser1.sub f +tests/posix2.tests f +tests/posix2.right f +tests/posixexp.tests f +tests/posixexp.right f +tests/posixexp1.sub f +tests/posixexp2.sub f +tests/posixexp2.tests f +tests/posixexp2.right f +tests/posixpat.tests f +tests/posixpat.right f +tests/posixpipe.tests f +tests/posixpipe.right f +tests/prec.right f +tests/precedence f +tests/printf.tests f +tests/printf.right f +tests/printf1.sub f +tests/printf2.sub f +tests/printf3.sub f +tests/printf4.sub f +tests/quote.tests f +tests/quote.right f +tests/quote1.sub f +tests/quote2.sub f +tests/read.tests f +tests/read.right f +tests/read1.sub f +tests/read2.sub f +tests/read3.sub f +tests/read4.sub f +tests/read5.sub f +tests/read6.sub f +tests/redir.tests f +tests/redir.right f +tests/redir1.sub f +tests/redir2.sub f +tests/redir3.sub f +tests/redir3.in1 f +tests/redir3.in2 f +tests/redir4.sub f +tests/redir4.in1 f +tests/redir5.sub f +tests/redir6.sub f +tests/redir7.sub f +tests/redir8.sub f +tests/redir9.sub f +tests/redir10.sub f +tests/rhs-exp.tests f +tests/rhs-exp.right f +tests/rhs-exp1.sub f +tests/rsh.tests f +tests/rsh.right f +tests/run-all f +tests/run-minimal f +tests/run-alias f +tests/run-appendop f +tests/run-arith-for f +tests/run-arith f +tests/run-array f +tests/run-array2 f +tests/run-assoc f +tests/run-braces f +tests/run-builtins f +tests/run-case f +tests/run-casemod f +tests/run-comsub f +tests/run-comsub-eof f +tests/run-comsub-posix f +tests/run-cond f +tests/run-coproc f +tests/run-cprint f +tests/run-dbg-support f +tests/run-dbg-support2 f +tests/run-dirstack f +tests/run-dollars f +tests/run-errors f +tests/run-execscript f +tests/run-exp-tests f +tests/run-exportfunc f +tests/run-extglob f +tests/run-extglob2 f +tests/run-extglob3 f +tests/run-func f +tests/run-getopts f +tests/run-glob-test f +tests/run-globstar f +tests/run-heredoc f +tests/run-herestr f +tests/run-histexpand f +tests/run-history f +tests/run-ifs f +tests/run-ifs-posix f +tests/run-input-test f +tests/run-intl f +tests/run-iquote f +tests/run-invert f +tests/run-jobs f +tests/run-lastpipe f +tests/run-mapfile f +tests/run-more-exp f +tests/run-nameref f +tests/run-new-exp f +tests/run-nquote f +tests/run-nquote1 f +tests/run-nquote2 f +tests/run-nquote3 f +tests/run-nquote4 f +tests/run-nquote5 f +tests/run-parser f +tests/run-posix2 f +tests/run-posixexp f +tests/run-posixexp2 f +tests/run-posixpat f +tests/run-posixpipe f +tests/run-precedence f +tests/run-printf f +tests/run-quote f +tests/run-read f +tests/run-redir f +tests/run-rhs-exp f +tests/run-rsh f +tests/run-set-e f +tests/run-set-x f +tests/run-shopt f +tests/run-strip f +tests/run-test f +tests/run-tilde f +tests/run-tilde2 f +tests/run-trap f +tests/run-type f +tests/run-varenv f +tests/run-vredir f +tests/set-e.tests f +tests/set-e1.sub f +tests/set-e2.sub f +tests/set-e3.sub f +tests/set-e3a.sub f +tests/set-e.right f +tests/set-x.tests f +tests/set-x1.sub f +tests/set-x.right f +tests/shopt.tests f +tests/shopt.right f +tests/strip.tests f +tests/strip.right f +tests/test.tests f +tests/test.right f +tests/tilde.tests f +tests/tilde.right f +tests/tilde2.tests f +tests/tilde2.right f +tests/trap.tests f +tests/trap.right f +tests/trap1.sub f 755 +tests/trap2.sub f 755 +tests/trap2a.sub f 755 +tests/trap3.sub f +tests/trap4.sub f +tests/trap5.sub f +tests/type.tests f +tests/type.right f +tests/type1.sub f +tests/type2.sub f +tests/type3.sub f +tests/type4.sub f +tests/unicode1.sub f +tests/unicode2.sub f +tests/unicode3.sub f +tests/varenv.right f +tests/varenv.sh f +tests/varenv1.sub f +tests/varenv2.sub f +tests/varenv3.sub f +tests/varenv4.sub f +tests/varenv5.sub f +tests/varenv6.sub f +tests/version f +tests/version.mini f +tests/vredir.tests f +tests/vredir.right f +tests/vredir1.sub f +tests/vredir2.sub f +tests/vredir3.sub f +tests/vredir4.sub f +tests/vredir5.sub f +tests/vredir6.sub f +tests/misc/dev-tcp.tests f +tests/misc/perf-script f +tests/misc/perftest f +tests/misc/read-nchars.tests f +tests/misc/redir-t2.sh f +tests/misc/run-r2.sh f +tests/misc/sigint-1.sh f +tests/misc/sigint-2.sh f +tests/misc/sigint-3.sh f +tests/misc/sigint-4.sh f +tests/misc/test-minus-e.1 f +tests/misc/test-minus-e.2 f +tests/misc/wait-bg.tests f +#examples/scripts.v2/PERMISSION f +#examples/scripts.v2/README f +#examples/scripts.v2/arc2tarz f +#examples/scripts.v2/bashrand f +#examples/scripts.v2/cal2day.bash f +#examples/scripts.v2/cdhist.bash f +#examples/scripts.v2/corename f +#examples/scripts.v2/fman f +#examples/scripts.v2/frcp f +#examples/scripts.v2/lowercase f +#examples/scripts.v2/ncp f +#examples/scripts.v2/newext f +#examples/scripts.v2/nmv f +#examples/scripts.v2/pages f +#examples/scripts.v2/pf f +#examples/scripts.v2/ren f +#examples/scripts.v2/rename f +#examples/scripts.v2/repeat f +#examples/scripts.v2/untar f +#examples/scripts.v2/uudec f +#examples/scripts.v2/uuenc f +#examples/scripts.v2/vtree f +#examples/scripts.v2/where f +#examples/scripts.v2/pmtop f +#examples/scripts.v2/shprof f +#examples/scripts.noah/PERMISSION f +#examples/scripts.noah/README f +#examples/scripts.noah/aref.bash f +#examples/scripts.noah/bash.sub.bash f +#examples/scripts.noah/bash_version.bash f +#examples/scripts.noah/meta.bash f +#examples/scripts.noah/mktmp.bash f +#examples/scripts.noah/number.bash f +#examples/scripts.noah/prompt.bash f +#examples/scripts.noah/remap_keys.bash f +#examples/scripts.noah/require.bash f +#examples/scripts.noah/send_mail.bash f +#examples/scripts.noah/shcat.bash f +#examples/scripts.noah/source.bash f +#examples/scripts.noah/string.bash f +#examples/scripts.noah/stty.bash f +#examples/scripts.noah/y_or_n_p.bash f diff --git a/cross-build/cygwin32.cache.old b/cross-build/cygwin32.cache.old new file mode 100644 index 000000000..640390fbf --- /dev/null +++ b/cross-build/cygwin32.cache.old @@ -0,0 +1,42 @@ +# This file is a shell script that caches the results of configure +# tests for CYGWIN32 so they don't need to be done when cross-compiling. + +# AC_FUNC_GETPGRP should also define GETPGRP_VOID +ac_cv_func_getpgrp_void=${ac_cv_func_getpgrp_void='yes'} +# AC_FUNC_SETVBUF_REVERSED should not define anything else +ac_cv_func_setvbuf_reversed=${ac_cv_func_setvbuf_reversed='no'} +# on CYGWIN32, system calls do not restart +ac_cv_sys_restartable_syscalls=${ac_cv_sys_restartable_syscalls='no'} +bash_cv_sys_restartable_syscalls=${bash_cv_sys_restartable_syscalls='no'} + +# these may be necessary, but they are currently commented out +#ac_cv_c_bigendian=${ac_cv_c_bigendian='no'} +ac_cv_sizeof_char_p=${ac_cv_sizeof_char_p='4'} +ac_cv_sizeof_int=${ac_cv_sizeof_int='4'} +ac_cv_sizeof_long=${ac_cv_sizeof_long='4'} +ac_cv_sizeof_double=${ac_cv_sizeof_double='8'} + +bash_cv_dup2_broken=${bash_cv_dup2_broken='no'} +bash_cv_pgrp_pipe=${bash_cv_pgrp_pipe='no'} +bash_cv_type_rlimit=${bash_cv_type_rlimit='long'} +bash_cv_decl_under_sys_siglist=${bash_cv_decl_under_sys_siglist='no'} +bash_cv_under_sys_siglist=${bash_cv_under_sys_siglist='no'} +bash_cv_sys_siglist=${bash_cv_sys_siglist='no'} +bash_cv_opendir_not_robust=${bash_cv_opendir_not_robust='no'} +bash_cv_getenv_redef=${bash_cv_getenv_redef='yes'} +bash_cv_printf_declared=${bash_cv_printf_declared='yes'} +bash_cv_ulimit_maxfds=${bash_cv_ulimit_maxfds='no'} +bash_cv_getcwd_calls_popen=${bash_cv_getcwd_calls_popen='no'} +bash_cv_must_reinstall_sighandlers=${bash_cv_must_reinstall_sighandlers='no'} +bash_cv_job_control_missing=${bash_cv_job_control_missing='present'} +bash_cv_sys_named_pipes=${bash_cv_sys_named_pipes='missing'} +bash_cv_func_sigsetjmp=${bash_cv_func_sigsetjmp='missing'} +bash_cv_mail_dir=${bash_cv_mail_dir='unknown'} +bash_cv_func_strcoll_broken=${bash_cv_func_strcoll_broken='no'} + +bash_cv_type_int32_t=${bash_cv_type_int32_t='int'} +bash_cv_type_u_int32_t=${bash_cv_type_u_int32_t='int'} + +ac_cv_type_bits64_t=${ac_cv_type_bits64_t='no'} + +# end of cross-build/cygwin32.cache diff --git a/doc/FAQ.orig b/doc/FAQ.orig new file mode 100644 index 000000000..1cff3c8ef --- /dev/null +++ b/doc/FAQ.orig @@ -0,0 +1,1745 @@ +This is the Bash FAQ, version 3.24, for Bash version 2.05b. + +This document contains a set of frequently-asked questions concerning +Bash, the GNU Bourne-Again Shell. Bash is a freely-available command +interpreter with advanced features for both interactive use and shell +programming. + +Another good source of basic information about shells is the collection +of FAQ articles periodically posted to comp.unix.shell. + +Questions and comments concerning this document should be sent to +chet@po.cwru.edu. + +This document is available for anonymous FTP with the URL + +ftp://ftp.cwru.edu/pub/bash/FAQ + +The Bash home page is http://cnswww.cns.cwru.edu/~chet/bash/bashtop.html + +---------- +Contents: + +Section A: The Basics + +A1) What is it? +A2) What's the latest version? +A3) Where can I get it? +A4) On what machines will bash run? +A5) Will bash run on operating systems other than Unix? +A6) How can I build bash with gcc? +A7) How can I make bash my login shell? +A8) I just changed my login shell to bash, and now I can't FTP into my + machine. Why not? +A9) What's the `POSIX 1003.2 standard'? +A10) What is the bash `posix mode'? + +Section B: The latest version + +B1) What's new in version 2.05b? +B2) Are there any user-visible incompatibilities between bash-2.05b and + bash-1.14.7? + +Section C: Differences from other Unix shells + +C1) How does bash differ from sh, the Bourne shell? +C2) How does bash differ from the Korn shell, version ksh88? +C3) Which new features in ksh-93 are not in bash, and which are? + +Section D: Why does bash do some things differently than other Unix shells? + +D1) Why does bash run a different version of `command' than + `which command' says it will? +D2) Why doesn't bash treat brace expansions exactly like csh? +D3) Why doesn't bash have csh variable modifiers? +D4) How can I make my csh aliases work when I convert to bash? +D5) How can I pipe standard output and standard error from one command to + another, like csh does with `|&'? +D6) Now that I've converted from ksh to bash, are there equivalents to + ksh features like autoloaded functions and the `whence' command? + +Section E: Why does bash do certain things the way it does? + +E1) Why is the bash builtin `test' slightly different from /bin/test? +E2) Why does bash sometimes say `Broken pipe'? +E3) When I have terminal escape sequences in my prompt, why does bash + wrap lines at the wrong column? +E4) If I pipe the output of a command into `read variable', why doesn't + the output show up in $variable when the read command finishes? +E5) I have a bunch of shell scripts that use backslash-escaped characters + in arguments to `echo'. Bash doesn't interpret these characters. Why + not, and how can I make it understand them? +E6) Why doesn't a while or for loop get suspended when I type ^Z? +E7) What about empty for loops in Makefiles? +E8) Why does the arithmetic evaluation code complain about `08'? +E9) Why does the pattern matching expression [A-Z]* match files beginning + with every letter except `z'? +E10) Why does `cd //' leave $PWD as `//'? +E11) If I resize my xterm while another program is running, why doesn't bash + notice the change? + +Section F: Things to watch out for on certain Unix versions + +F1) Why can't I use command line editing in my `cmdtool'? +F2) I built bash on Solaris 2. Why do globbing expansions and filename + completion chop off the first few characters of each filename? +F3) Why does bash dump core after I interrupt username completion or + `~user' tilde expansion on a machine running NIS? +F4) I'm running SVR4.2. Why is the line erased every time I type `@'? +F5) Why does bash report syntax errors when my C News scripts use a + redirection before a subshell command? +F6) Why can't I use vi-mode editing on Red Hat Linux 6.1? +F7) Why do bash-2.05a and bash-2.05b fail to compile `printf.def' on + HP/UX 11.x? + +Section G: How can I get bash to do certain common things? + +G1) How can I get bash to read and display eight-bit characters? +G2) How do I write a function `x' to replace builtin command `x', but + still invoke the command from within the function? +G3) How can I find the value of a shell variable whose name is the value + of another shell variable? +G4) How can I make the bash `time' reserved word print timing output that + looks like the output from my system's /usr/bin/time? +G5) How do I get the current directory into my prompt? +G6) How can I rename "*.foo" to "*.bar"? +G7) How can I translate a filename from uppercase to lowercase? +G8) How can I write a filename expansion (globbing) pattern that will match + all files in the current directory except "." and ".."? + +Section H: Where do I go from here? + +H1) How do I report bugs in bash, and where should I look for fixes and + advice? +H2) What kind of bash documentation is there? +H3) What's coming in future versions? +H4) What's on the bash `wish list'? +H5) When will the next release appear? + +---------- +Section A: The Basics + +A1) What is it? + +Bash is a Unix command interpreter (shell). It is an implementation of +the Posix 1003.2 shell standard, and resembles the Korn and System V +shells. + +Bash contains a number of enhancements over those shells, both +for interactive use and shell programming. Features geared +toward interactive use include command line editing, command +history, job control, aliases, and prompt expansion. Programming +features include additional variable expansions, shell +arithmetic, and a number of variables and options to control +shell behavior. + +Bash was originally written by Brian Fox of the Free Software +Foundation. The current developer and maintainer is Chet Ramey +of Case Western Reserve University. + +A2) What's the latest version? + +The latest version is 2.05b, first made available on Wednesday, 17 +July, 2002. + +A3) Where can I get it? + +Bash is the GNU project's shell, and so is available from the +master GNU archive site, ftp.gnu.org, and its mirrors. The +latest version is also available for FTP from ftp.cwru.edu. +The following URLs tell how to get version 2.05b: + +ftp://ftp.gnu.org/pub/gnu/bash/bash-2.05b.tar.gz +ftp://ftp.cwru.edu/pub/bash/bash-2.05b.tar.gz + +Formatted versions of the documentation are available with the URLs: + +ftp://ftp.gnu.org/pub/gnu/bash/bash-doc-2.05b.tar.gz +ftp://ftp.cwru.edu/pub/bash/bash-doc-2.05b.tar.gz + +A4) On what machines will bash run? + +Bash has been ported to nearly every version of UNIX. All you +should have to do to build it on a machine for which a port +exists is to type `configure' and then `make'. The build process +will attempt to discover the version of UNIX you have and tailor +itself accordingly, using a script created by GNU autoconf. + +More information appears in the file `INSTALL' in the distribution. + +The Bash web page (http://cnswww.cns.cwru.edu/~chet/bash/bashtop.html) +explains how to obtain binary versions of bash for most of the major +commercial Unix systems. + +A5) Will bash run on operating systems other than Unix? + +Configuration specifics for Unix-like systems such as QNX and +LynxOS are included in the distribution. Bash-2.05 and later +versions should compile and run on Minix 2.0 (patches were +contributed), but I don't believe anyone has built bash-2.x on +earlier Minix versions yet. + +Bash has been ported to versions of Windows implementing the Win32 +programming interface. This includes Windows 95 and Windows NT. +The port was done by Cygnus Solutions as part of their CYGWIN +project. For more information about the project, look at the URLs + +http://www.cygwin.com/ +http://sourceware.cygnus.com/cygwin + +Cygnus originally ported bash-1.14.7, and that port was part of their +early GNU-Win32 (the original name) releases. Cygnus has also done a +port of bash-2.05 to the CYGWIN environment, and it is available as +part of their current release. + +Bash-2.05b should require no local Cygnus changes to build and run under +CYGWIN. + +The Cygnus port works only on Intel machines. There is a port of bash +(I don't know which version) to the alpha/NT environment available from + +ftp://ftp.gnustep.org//pub/win32/bash-alpha-nt-1.01.tar.gz + +DJ Delorie has a port of bash-2.x which runs under MS-DOS, as part +of the DJGPP project. For more information on the project, see + +http://www.delorie.com/djgpp/ + +I have been told that the original DJGPP port was done by Daisuke Aoyama. + +Mark Elbrecht has sent me notice that bash-2.04 +is available for DJGPP V2. The files are available as: + +ftp://ftp.simtel.net/pub/simtelnet/gnu/djgpp/v2gnu/bsh204b.zip binary +ftp://ftp.simtel.net/pub/simtelnet/gnu/djgpp/v2gnu/bsh204d.zip documentation +ftp://ftp.simtel.net/pub/simtelnet/gnu/djgpp/v2gnu/bsh204s.zip source + +Mark has begun to work with bash-2.05, but I don't know the status. + +Ports of bash-1.12 and bash-2.0 are available for OS/2 from + +ftp://hobbes.nmsu.edu/pub/os2/util/shell/bash_112.zip +ftp://hobbes.nmsu.edu/pub/os2/util/shell/bash-2.0(253).zip + +I haven't looked at either, but the second appears to be a binary-only +distribution. Beware. + +I have received word that Bash (I'm not sure which version, but I +believe that it's at least bash-2.02.1) is the standard shell on +BeOS. + +A6) How can I build bash with gcc? + +Bash configures to use gcc by default if it is available. Read the +file INSTALL in the distribution for more information. + +A7) How can I make bash my login shell? + +Some machines let you use `chsh' to change your login shell. Other +systems use `passwd -s' or `passwd -e'. If one of these works for +you, that's all you need. Note that many systems require the full +pathname to a shell to appear in /etc/shells before you can make it +your login shell. For this, you may need the assistance of your +friendly local system administrator. + +If you cannot do this, you can still use bash as your login shell, but +you need to perform some tricks. The basic idea is to add a command +to your login shell's startup file to replace your login shell with +bash. + +For example, if your login shell is csh or tcsh, and you have installed +bash in /usr/gnu/bin/bash, add the following line to ~/.login: + + if ( -f /usr/gnu/bin/bash ) exec /usr/gnu/bin/bash --login + +(the `--login' tells bash that it is a login shell). + +It's not a good idea to put this command into ~/.cshrc, because every +csh you run without the `-f' option, even ones started to run csh scripts, +reads that file. If you must put the command in ~/.cshrc, use something +like + + if ( $?prompt ) exec /usr/gnu/bin/bash --login + +to ensure that bash is exec'd only when the csh is interactive. + +If your login shell is sh or ksh, you have to do two things. + +First, create an empty file in your home directory named `.bash_profile'. +The existence of this file will prevent the exec'd bash from trying to +read ~/.profile, and re-execing itself over and over again. ~/.bash_profile +is the first file bash tries to read initialization commands from when +it is invoked as a login shell. + +Next, add a line similar to the above to ~/.profile: + + [ -f /usr/gnu/bin/bash ] && [ -x /usr/gnu/bin/bash ] && \ + exec /usr/gnu/bin/bash --login + +This will cause login shells to replace themselves with bash running as +a login shell. Once you have this working, you can copy your initialization +code from ~/.profile to ~/.bash_profile. + +I have received word that the recipe supplied above is insufficient for +machines running CDE. CDE has a maze of twisty little startup files, all +slightly different. + +If you cannot change your login shell in the password file to bash, you +will have to (apparently) live with CDE using the shell in the password +file to run its startup scripts. If you have changed your shell to bash, +there is code in the CDE startup files (on Solaris, at least) that attempts +to do the right thing. It is, however, often broken, and may require that +you use the $BASH_ENV trick described below. + +`dtterm' claims to use $SHELL as the default program to start, so if you +can change $SHELL in the CDE startup files, you should be able to use bash +in your terminal windows. + +Setting DTSOURCEPROFILE in ~/.dtprofile will cause the `Xsession' program +to read your login shell's startup files. You may be able to use bash for +the rest of the CDE programs by setting SHELL to bash in ~/.dtprofile as +well, but I have not tried this. + +You can use the above `exec' recipe to start bash when not logging in with +CDE by testing the value of the DT variable: + + if [ -n "$DT" ]; then + [ -f /usr/gnu/bin/bash ] && exec /usr/gnu/bin/bash --login + fi + +If CDE starts its shells non-interactively during login, the login shell +startup files (~/.profile, ~/.bash_profile) will not be sourced at login. +To get around this problem, append a line similar to the following to your +~/.dtprofile: + + BASH_ENV=${HOME}/.bash_profile ; export BASH_ENV + +and add the following line to the beginning of ~/.bash_profile: + + unset BASH_ENV + +A8) I just changed my login shell to bash, and now I can't FTP into my + machine. Why not? + +You must add the full pathname to bash to the file /etc/shells. As +noted in the answer to the previous question, many systems require +this before you can make bash your login shell. + +Most versions of ftpd use this file to prohibit `special' users +such as `uucp' and `news' from using FTP. + +A9) What's the `POSIX 1003.2 standard'? + +POSIX is a name originally coined by Richard Stallman for a +family of open system standards based on UNIX. There are a +number of aspects of UNIX under consideration for +standardization, from the basic system services at the system +call and C library level to applications and tools to system +administration and management. Each area of standardization is +assigned to a working group in the 1003 series. + +The POSIX Shell and Utilities standard has been developed by IEEE +Working Group 1003.2 (POSIX.2). It concentrates on the command +interpreter interface and utility programs commonly executed from +the command line or by other programs. An initial version of the +standard has been approved and published by the IEEE, and work is +currently underway to update it. + +Bash is concerned with the aspects of the shell's behavior +defined by POSIX.2. The shell command language has of course +been standardized, including the basic flow control and program +execution constructs, I/O redirection and pipelining, argument +handling, variable expansion, and quoting. + +The `special' builtins, which must be implemented as part of the +shell to provide the desired functionality, are specified as +being part of the shell; examples of these are `eval' and +`export'. Other utilities appear in the sections of POSIX.2 not +devoted to the shell which are commonly (and in some cases must +be) implemented as builtin commands, such as `read' and `test'. +POSIX.2 also specifies aspects of the shell's interactive +behavior as part of the UPE, including job control and command +line editing. Only vi-style line editing commands have been +standardized; emacs editing commands were left out due to +objections. + +The Open Group has made an older version of its Single Unix +Specification (version 2), which is very similar to POSIX.2, +available on the web at + +http://www.opengroup.org/onlinepubs/007908799/ + +The Single Unix Specification, version 3, is available on the web at + +http://www.opengroup.org/onlinepubs/007904975/ + +A10) What is the bash `posix mode'? + +Although bash is an implementation of the POSIX.2 shell +specification, there are areas where the bash default behavior +differs from that spec. The bash `posix mode' changes the bash +behavior in these areas so that it obeys the spec more closely. + +Posix mode is entered by starting bash with the --posix or +'-o posix' option or executing `set -o posix' after bash is running. + +The specific aspects of bash which change when posix mode is +active are listed in the file POSIX in the bash distribution. +They are also listed in a section in the Bash Reference Manual +(from which that file is generated). + +Section B: The latest version + +B1) What's new in version 2.05b? + +The raison d'etre for bash-2.05b is to make a second intermediate +release containing the first of the new features to be available +in bash-3.0 and get feedback on those features before proceeding. +The major new feature is multibyte character support in both Bash +and Readline. + +Bash-2.05b contains the following new features (see the manual page for +complete descriptions and the CHANGES and NEWS files in the bash-2.05b +distribution): + +o support for multibyte characters has been added to both bash and readline + +o the DEBUG trap is now run *before* simple commands, ((...)) commands, + [[...]] conditional commands, and for ((...)) loops + +o the shell now performs arithmetic in the largest integer size the machine + supports (intmax_t) + +o there is a new \D{...} prompt expansion; passes the `...' to strftime(3) + and inserts the result into the expanded prompt + +o there is a new `here-string' redirection operator: <<< word + +o when displaying variables, function attributes and definitions are shown + separately, allowing them to be re-used as input (attempting to re-use + the old output would result in syntax errors). + +o `read' has a new `-u fd' option to read from a specified file descriptor + +o the bash debugger in examples/bashdb has been modified to work with the + new DEBUG trap semantics, the command set has been made more gdb-like, + and the changes to $LINENO make debugging functions work better + +o the expansion of $LINENO inside a shell function is only relative to the + function start if the shell is interactive -- if the shell is running a + script, $LINENO expands to the line number in the script. This is as + POSIX-2001 requires + + +A short feature history dating from Bash-2.0: + +Bash-2.05a introduced the following new features: + +o The `printf' builtin has undergone major work + +o There is a new read-only `shopt' option: login_shell, which is set by + login shells and unset otherwise + +o New `\A' prompt string escape sequence; expanding to time in 24-hour + HH:MM format + +o New `-A group/-g' option to complete and compgen; goes group name + completion + +o New [+-]O invocation option to set and unset `shopt' options at startup + +o ksh-like `ERR' trap + +o `for' loops now allow empty word lists after the `in' reserved word + +o new `hard' and `soft' arguments for the `ulimit' builtin + +o Readline can be configured to place the user at the same point on the line + when retrieving commands from the history list + +o Readline can be configured to skip `hidden' files (filenames with a leading + `.' on Unix) when performing completion + +Bash-2.05 introduced the following new features: + +o This version has once again reverted to using locales and strcoll(3) when + processing pattern matching bracket expressions, as POSIX requires. +o Added a new `--init-file' invocation argument as a synonym for `--rcfile', + per the new GNU coding standards. +o The /dev/tcp and /dev/udp redirections now accept service names as well as + port numbers. +o `complete' and `compgen' now take a `-o value' option, which controls some + of the aspects of that compspec. Valid values are: + + default - perform bash default completion if programmable + completion produces no matches + dirnames - perform directory name completion if programmable + completion produces no matches + filenames - tell readline that the compspec produces filenames, + so it can do things like append slashes to + directory names and suppress trailing spaces +o A new loadable builtin, realpath, which canonicalizes and expands symlinks + in pathname arguments. +o When `set' is called without options, it prints function defintions in a + way that allows them to be reused as input. This affects `declare' and + `declare -p' as well. This only happens when the shell is not in POSIX + mode, since POSIX.2 forbids this behavior. + +Bash-2.04 introduced the following new features: + +o Programmable word completion with the new `complete' and `compgen' builtins; + examples are provided in examples/complete/complete-examples +o `history' has a new `-d' option to delete a history entry +o `bind' has a new `-x' option to bind key sequences to shell commands +o The prompt expansion code has new `\j' and `\l' escape sequences +o The `no_empty_cmd_completion' shell option, if enabled, inhibits + command completion when TAB is typed on an empty line +o `help' has a new `-s' option to print a usage synopsis +o New arithmetic operators: var++, var--, ++var, --var, expr1,expr2 (comma) +o New ksh93-style arithmetic for command: + for ((expr1 ; expr2; expr3 )); do list; done +o `read' has new options: `-t', `-n', `-d', `-s' +o The redirection code handles several filenames specially: /dev/fd/N, + /dev/stdin, /dev/stdout, /dev/stderr +o The redirection code now recognizes /dev/tcp/HOST/PORT and + /dev/udp/HOST/PORT and tries to open a TCP or UDP socket, respectively, + to the specified port on the specified host +o The ${!prefix*} expansion has been implemented +o A new FUNCNAME variable, which expands to the name of a currently-executing + function +o The GROUPS variable is no longer readonly +o A new shopt `xpg_echo' variable, to control the behavior of echo with + respect to backslash-escape sequences at runtime +o The NON_INTERACTIVE_LOGIN_SHELLS #define has returned + +The version of Readline released with Bash-2.04, Readline-4.1, had several +new features as well: + +o Parentheses matching is always compiled into readline, and controllable + with the new `blink-matching-paren' variable +o The history-search-forward and history-search-backward functions now leave + point at the end of the line when the search string is empty, like + reverse-search-history, and forward-search-history +o A new function for applications: rl_on_new_line_with_prompt() +o New variables for applications: rl_already_prompted, and rl_gnu_readline_p + + +Bash-2.03 had very few new features, in keeping with the convention +that odd-numbered releases provide mainly bug fixes. A number of new +features were added to Readline, mostly at the request of the Cygnus +folks. + +A new shopt option, `restricted_shell', so that startup files can test + whether or not the shell was started in restricted mode +Filename generation is now performed on the words between ( and ) in + compound array assignments (this is really a bug fix) +OLDPWD is now auto-exported, as POSIX.2 requires +ENV and BASH_ENV are read-only variables in a restricted shell +Bash may now be linked against an already-installed Readline library, + as long as the Readline library is version 4 or newer +All shells begun with the `--login' option will source the login shell + startup files, even if the shell is not interactive + +There were lots of changes to the version of the Readline library released +along with Bash-2.03. For a complete list of the changes, read the file +CHANGES in the Bash-2.03 distribution. + +Bash-2.02 contained the following new features: + +a new version of malloc (based on the old GNU malloc code in previous + bash versions) that is more page-oriented, more conservative + with memory usage, does not `orphan' large blocks when they + are freed, is usable on 64-bit machines, and has allocation + checking turned on unconditionally +POSIX.2-style globbing character classes ([:alpha:], [:alnum:], etc.) +POSIX.2-style globbing equivalence classes +POSIX.2-style globbing collating symbols +the ksh [[...]] extended conditional command +the ksh egrep-style extended pattern matching operators +a new `printf' builtin +the ksh-like $(, &>, >|, <<<, [n]<&word-, [n]>&word- + prompt string special char translation and variable expansion + auto-export of variables in initial environment + command search finds functions before builtins + bash return builtin will exit a file sourced with `.' + builtins: cd -/-L/-P, exec -l/-c/-a, echo -e/-E, hash -d/-l/-p/-t. + export -n/-f/-p/name=value, pwd -L/-P, + read -e/-p/-a/-t/-n/-d/-s/-u, + readonly -a/-f/name=value, trap -l, set +o, + set -b/-m/-o option/-h/-p/-B/-C/-H/-P, + unset -f/-v, ulimit -m/-p/-u, + type -a/-p/-t/-f/-P, suspend -f, kill -n, + test -o optname/s1 == s2/s1 < s2/s1 > s2/-nt/-ot/-ef/-O/-G/-S + bash reads ~/.bashrc for interactive shells, $ENV for non-interactive + bash restricted shell mode is more extensive + bash allows functions and variables with the same name + brace expansion + tilde expansion + arithmetic expansion with $((...)) and `let' builtin + the `[[...]]' extended conditional command + process substitution + aliases and alias/unalias builtins + local variables in functions and `local' builtin + readline and command-line editing with programmable completion + command history and history/fc builtins + csh-like history expansion + other new bash builtins: bind, command, compgen, complete, builtin, + declare/typeset, dirs, enable, fc, help, + history, logout, popd, pushd, disown, shopt, + printf + exported functions + filename generation when using output redirection (command >a*) + POSIX.2-style globbing character classes + POSIX.2-style globbing equivalence classes + POSIX.2-style globbing collating symbols + egrep-like extended pattern matching operators + case-insensitive pattern matching and globbing + variable assignments preceding commands affect only that command, + even for builtins and functions + posix mode + redirection to /dev/fd/N, /dev/stdin, /dev/stdout, /dev/stderr, + /dev/tcp/host/port, /dev/udp/host/port + +Things sh has that bash does not: + uses variable SHACCT to do shell accounting + includes `stop' builtin (bash can use alias stop='kill -s STOP') + `newgrp' builtin + turns on job control if called as `jsh' + $TIMEOUT (like bash $TMOUT) + `^' is a synonym for `|' + new SVR4.2 sh builtins: mldmode, priv + +Implementation differences: + redirection to/from compound commands causes sh to create a subshell + bash does not allow unbalanced quotes; sh silently inserts them at EOF + bash does not mess with signal 11 + sh sets (euid, egid) to (uid, gid) if -p not supplied and uid < 100 + bash splits only the results of expansions on IFS, using POSIX.2 + field splitting rules; sh splits all words on IFS + sh does not allow MAILCHECK to be unset (?) + sh does not allow traps on SIGALRM or SIGCHLD + bash allows multiple option arguments when invoked (e.g. -x -v); + sh allows only a single option argument (`sh -x -v' attempts + to open a file named `-v', and, on SunOS 4.1.4, dumps core. + On Solaris 2.4 and earlier versions, sh goes into an infinite + loop.) + sh exits a script if any builtin fails; bash exits only if one of + the POSIX.2 `special' builtins fails + +C2) How does bash differ from the Korn shell, version ksh88? + +Things bash has or uses that ksh88 does not: + long invocation options + [-+]O invocation option + -l invocation option + `!' reserved word + arithmetic for command: for ((expr1 ; expr2; expr3 )); do list; done + arithmetic in largest machine-supported size (intmax_t) + posix mode and posix conformance + command hashing + tilde expansion for assignment statements that look like $PATH + process substitution with named pipes if /dev/fd is not available + the ${!param} indirect parameter expansion operator + the ${!param*} prefix expansion operator + the ${param:offset[:length]} parameter substring operator + the ${param/pat[/string]} parameter pattern substitution operator + variables: BASH, BASH_VERSION, BASH_VERSINFO, UID, EUID, SHLVL, + TIMEFORMAT, HISTCMD, HOSTTYPE, OSTYPE, MACHTYPE, + HISTFILESIZE, HISTIGNORE, HISTCONTROL, PROMPT_COMMAND, + IGNOREEOF, FIGNORE, INPUTRC, HOSTFILE, DIRSTACK, + PIPESTATUS, HOSTNAME, OPTERR, SHELLOPTS, GLOBIGNORE, + GROUPS, FUNCNAME, histchars, auto_resume + prompt expansion with backslash escapes and command substitution + redirection: &> (stdout and stderr), <<<, [n]<&word-, [n]>&word- + more extensive and extensible editing and programmable completion + builtins: bind, builtin, command, declare, dirs, echo -e/-E, enable, + exec -l/-c/-a, fc -s, export -n/-f/-p, hash, help, history, + jobs -x/-r/-s, kill -s/-n/-l, local, logout, popd, pushd, + read -e/-p/-a/-t/-n/-d/-s, readonly -a/-n/-f/-p, + set -o braceexpand/-o histexpand/-o interactive-comments/ + -o notify/-o physical/-o posix/-o hashall/-o onecmd/ + -h/-B/-C/-b/-H/-P, set +o, suspend, trap -l, type, + typeset -a/-F/-p, ulimit -u, umask -S, alias -p, shopt, + disown, printf, complete, compgen + `!' csh-style history expansion + POSIX.2-style globbing character classes + POSIX.2-style globbing equivalence classes + POSIX.2-style globbing collating symbols + egrep-like extended pattern matching operators + case-insensitive pattern matching and globbing + `**' arithmetic operator to do exponentiation + redirection to /dev/fd/N, /dev/stdin, /dev/stdout, /dev/stderr + arrays of unlimited size + TMOUT is default timeout for `read' and `select' + +Things ksh88 has or uses that bash does not: + tracked aliases (alias -t) + variables: ERRNO, FPATH, EDITOR, VISUAL + co-processes (|&, >&p, <&p) + weirdly-scoped functions + typeset +f to list all function names without definitions + text of command history kept in a file, not memory + builtins: alias -x, cd old new, fc -e -, newgrp, print, + read -p/-s/var?prompt, set -A/-o gmacs/ + -o bgnice/-o markdirs/-o nolog/-o trackall/-o viraw/-s, + typeset -H/-L/-R/-Z/-A/-ft/-fu/-fx/-l/-u/-t, whence + using environment to pass attributes of exported variables + arithmetic evaluation done on arguments to some builtins + reads .profile from $PWD when invoked as login shell + +Implementation differences: + ksh runs last command of a pipeline in parent shell context + bash has brace expansion by default (ksh88 compile-time option) + bash has fixed startup file for all interactive shells; ksh reads $ENV + bash has exported functions + bash command search finds functions before builtins + bash waits for all commands in pipeline to exit before returning status + emacs-mode editing has some slightly different key bindings + +C3) Which new features in ksh-93 are not in bash, and which are? + +New things in ksh-93 not in bash-2.05b: + associative arrays + floating point arithmetic and variables + math library functions + ${!name[sub]} name of subscript for associative array + `.' is allowed in variable names to create a hierarchical namespace + more extensive compound assignment syntax + discipline functions + `sleep' and `getconf' builtins (bash has loadable versions) + typeset -n and `nameref' variables + KEYBD trap + variables: .sh.edchar, .sh.edmode, .sh.edcol, .sh.edtext, .sh.version, + .sh.name, .sh.subscript, .sh.value, .sh.match, HISTEDIT + backreferences in pattern matching (\N) + `&' operator in pattern lists for matching + print -f (bash uses printf) + `fc' has been renamed to `hist' + `.' can execute shell functions + exit statuses between 0 and 255 + set -o pipefail + `+=' variable assignment operator + FPATH and PATH mixing + getopts -a + -I invocation option + DEBUG trap now executed before each simple command, instead of after + printf %H, %P, %T, %Z modifiers, output base for %d + lexical scoping for local variables in `ksh' functions + no scoping for local variables in `POSIX' functions + +New things in ksh-93 present in bash-2.05b: + [n]<&word- and [n]>&word- redirections (combination dup and close) + for (( expr1; expr2; expr3 )) ; do list; done - arithmetic for command + ?:, ++, --, `expr1 , expr2' arithmetic operators + expansions: ${!param}, ${param:offset[:len]}, ${param/pat[/str]}, + ${!param*} + compound array assignment + the `!' reserved word + loadable builtins -- but ksh uses `builtin' while bash uses `enable' + `command', `builtin', `disown' builtins + new $'...' and $"..." quoting + FIGNORE (but bash uses GLOBIGNORE), HISTCMD + set -o notify/-C + changes to kill builtin + read -A (bash uses read -a) + read -t/-d + trap -p + exec -c/-a + `.' restores the positional parameters when it completes + POSIX.2 `test' + umask -S + unalias -a + command and arithmetic substitution performed on PS1, PS4, and ENV + command name completion + ENV processed only for interactive shells + +Section D: Why does bash do some things differently than other Unix shells? + +D1) Why does bash run a different version of `command' than + `which command' says it will? + +On many systems, `which' is actually a csh script that assumes +you're running csh. In tcsh, `which' and its cousin `where' +are builtins. On other Unix systems, `which' is a perl script +that uses the PATH environment variable. + +The csh script version reads the csh startup files from your +home directory and uses those to determine which `command' will +be invoked. Since bash doesn't use any of those startup files, +there's a good chance that your bash environment differs from +your csh environment. The bash `type' builtin does everything +`which' does, and will report correct results for the running +shell. If you're really wedded to the name `which', try adding +the following function definition to your .bashrc: + + which() + { + builtin type "$@" + } + +If you're moving from tcsh and would like to bring `where' along +as well, use this function: + + where() + { + builtin type -a "$@" + } + +D2) Why doesn't bash treat brace expansions exactly like csh? + +The only difference between bash and csh brace expansion is that +bash requires a brace expression to contain at least one unquoted +comma if it is to be expanded. Any brace-surrounded word not +containing an unquoted comma is left unchanged by the brace +expansion code. This affords the greatest degree of sh +compatibility. + +Bash, ksh, zsh, and pd-ksh all implement brace expansion this way. + +D3) Why doesn't bash have csh variable modifiers? + +Posix has specified a more powerful, albeit somewhat more cryptic, +mechanism cribbed from ksh, and bash implements it. + +${parameter%word} + Remove smallest suffix pattern. The WORD is expanded to produce + a pattern. It then expands to the value of PARAMETER, with the + smallest portion of the suffix matched by the pattern deleted. + + x=file.c + echo ${x%.c}.o + -->file.o + +${parameter%%word} + + Remove largest suffix pattern. The WORD is expanded to produce + a pattern. It then expands to the value of PARAMETER, with the + largest portion of the suffix matched by the pattern deleted. + + x=posix/src/std + echo ${x%%/*} + -->posix + +${parameter#word} + Remove smallest prefix pattern. The WORD is expanded to produce + a pattern. It then expands to the value of PARAMETER, with the + smallest portion of the prefix matched by the pattern deleted. + + x=$HOME/src/cmd + echo ${x#$HOME} + -->/src/cmd + +${parameter##word} + Remove largest prefix pattern. The WORD is expanded to produce + a pattern. It then expands to the value of PARAMETER, with the + largest portion of the prefix matched by the pattern deleted. + + x=/one/two/three + echo ${x##*/} + -->three + + +Given + a=/a/b/c/d + b=b.xxx + + csh bash result + --- ---- ------ + $a:h ${a%/*} /a/b/c + $a:t ${a##*/} d + $b:r ${b%.*} b + $b:e ${b##*.} xxx + + +D4) How can I make my csh aliases work when I convert to bash? + +Bash uses a different syntax to support aliases than csh does. +The details can be found in the documentation. We have provided +a shell script which does most of the work of conversion for you; +this script can be found in ./examples/misc/aliasconv.sh. Here is +how you use it: + +Start csh in the normal way for you. (e.g., `csh') + +Pipe the output of `alias' through `aliasconv.sh', saving the +results into `bash_aliases': + + alias | bash aliasconv.sh >bash_aliases + +Edit `bash_aliases', carefully reading through any created +functions. You will need to change the names of some csh specific +variables to the bash equivalents. The script converts $cwd to +$PWD, $term to $TERM, $home to $HOME, $user to $USER, and $prompt +to $PS1. You may also have to add quotes to avoid unwanted +expansion. + +For example, the csh alias: + + alias cd 'cd \!*; echo $cwd' + +is converted to the bash function: + + cd () { command cd "$@"; echo $PWD ; } + +The only thing that needs to be done is to quote $PWD: + + cd () { command cd "$@"; echo "$PWD" ; } + +Merge the edited file into your ~/.bashrc. + +There is an additional, more ambitious, script in +examples/misc/cshtobash that attempts to convert your entire csh +environment to its bash equivalent. This script can be run as +simply `cshtobash' to convert your normal interactive +environment, or as `cshtobash ~/.login' to convert your login +environment. + +D5) How can I pipe standard output and standard error from one command to + another, like csh does with `|&'? + +Use + command 2>&1 | command2 + +The key is to remember that piping is performed before redirection, so +file descriptor 1 points to the pipe when it is duplicated onto file +descriptor 2. + +D6) Now that I've converted from ksh to bash, are there equivalents to + ksh features like autoloaded functions and the `whence' command? + +There are features in ksh-88 and ksh-93 that do not have direct bash +equivalents. Most, however, can be emulated with very little trouble. + +ksh-88 feature Bash equivalent +-------------- --------------- +compiled-in aliases set up aliases in .bashrc; some ksh aliases are + bash builtins (hash, history, type) +coprocesses named pipe pairs (one for read, one for write) +typeset +f declare -F +cd, print, whence function substitutes in examples/functions/kshenv +autoloaded functions examples/functions/autoload is the same as typeset -fu +read var?prompt read -p prompt var + +ksh-93 feature Bash equivalent +-------------- --------------- +sleep, getconf Bash has loadable versions in examples/loadables +${.sh.version} $BASH_VERSION +print -f printf +hist alias hist=fc +$HISTEDIT $FCEDIT + +Section E: How can I get bash to do certain things, and why does bash do + things the way it does? + +E1) Why is the bash builtin `test' slightly different from /bin/test? + +The specific example used here is [ ! x -o x ], which is false. + +Bash's builtin `test' implements the Posix.2 spec, which can be +summarized as follows (the wording is due to David Korn): + +Here is the set of rules for processing test arguments. + + 0 Args: False + 1 Arg: True iff argument is not null. + 2 Args: If first arg is !, True iff second argument is null. + If first argument is unary, then true if unary test is true + Otherwise error. + 3 Args: If second argument is a binary operator, do binary test of $1 $3 + If first argument is !, negate two argument test of $2 $3 + If first argument is `(' and third argument is `)', do the + one-argument test of the second argument. + Otherwise error. + 4 Args: If first argument is !, negate three argument test of $2 $3 $4. + Otherwise unspecified + 5 or more Args: unspecified. (Historical shells would use their + current algorithm). + +The operators -a and -o are considered binary operators for the purpose +of the 3 Arg case. + +As you can see, the test becomes (not (x or x)), which is false. + +E2) Why does bash sometimes say `Broken pipe'? + +If a sequence of commands appears in a pipeline, and one of the +reading commands finishes before the writer has finished, the +writer receives a SIGPIPE signal. Many other shells special-case +SIGPIPE as an exit status in the pipeline and do not report it. +For example, in: + + ps -aux | head + +`head' can finish before `ps' writes all of its output, and ps +will try to write on a pipe without a reader. In that case, bash +will print `Broken pipe' to stderr when ps is killed by a +SIGPIPE. + +You can build a version of bash that will not report SIGPIPE errors +by uncommenting the definition of DONT_REPORT_SIGPIPE in the file +config-top.h. + +E3) When I have terminal escape sequences in my prompt, why does bash + wrap lines at the wrong column? + +Readline, the line editing library that bash uses, does not know +that the terminal escape sequences do not take up space on the +screen. The redisplay code assumes, unless told otherwise, that +each character in the prompt is a `printable' character that +takes up one character position on the screen. + +You can use the bash prompt expansion facility (see the PROMPTING +section in the manual page) to tell readline that sequences of +characters in the prompt strings take up no screen space. + +Use the \[ escape to begin a sequence of non-printing characters, +and the \] escape to signal the end of such a sequence. + +E4) If I pipe the output of a command into `read variable', why doesn't + the output show up in $variable when the read command finishes? + +This has to do with the parent-child relationship between Unix +processes. It affects all commands run in pipelines, not just +simple calls to `read'. For example, piping a command's output +into a `while' loop that repeatedly calls `read' will result in +the same behavior. + +Each element of a pipeline runs in a separate process, a child of +the shell running the pipeline. A subprocess cannot affect its +parent's environment. When the `read' command sets the variable +to the input, that variable is set only in the subshell, not the +parent shell. When the subshell exits, the value of the variable +is lost. + +Many pipelines that end with `read variable' can be converted +into command substitutions, which will capture the output of +a specified command. The output can then be assigned to a +variable: + + grep ^gnu /usr/lib/news/active | wc -l | read ngroup + +can be converted into + + ngroup=$(grep ^gnu /usr/lib/news/active | wc -l) + +This does not, unfortunately, work to split the text among +multiple variables, as read does when given multiple variable +arguments. If you need to do this, you can either use the +command substitution above to read the output into a variable +and chop up the variable using the bash pattern removal +expansion operators or use some variant of the following +approach. + +Say /usr/local/bin/ipaddr is the following shell script: + +#! /bin/sh +host `hostname` | awk '/address/ {print $NF}' + +Instead of using + + /usr/local/bin/ipaddr | read A B C D + +to break the local machine's IP address into separate octets, use + + OIFS="$IFS" + IFS=. + set -- $(/usr/local/bin/ipaddr) + IFS="$OIFS" + A="$1" B="$2" C="$3" D="$4" + +Beware, however, that this will change the shell's positional +parameters. If you need them, you should save them before doing +this. + +This is the general approach -- in most cases you will not need to +set $IFS to a different value. + +Some other user-supplied alternatives include: + +read A B C D << HERE + $(IFS=.; echo $(/usr/local/bin/ipaddr)) +HERE + +and, where process substitution is available, + +read A B C D < <(IFS=.; echo $(/usr/local/bin/ipaddr)) + +E5) I have a bunch of shell scripts that use backslash-escaped characters + in arguments to `echo'. Bash doesn't interpret these characters. Why + not, and how can I make it understand them? + +This is the behavior of echo on most Unix System V machines. + +The bash builtin `echo' is modeled after the 9th Edition +Research Unix version of `echo'. It does not interpret +backslash-escaped characters in its argument strings by default; +it requires the use of the -e option to enable the +interpretation. The System V echo provides no way to disable the +special characters; the bash echo has a -E option to disable +them. + +There is a configuration option that will make bash behave like +the System V echo and interpret things like `\t' by default. Run +configure with the --enable-xpg-echo-default option to turn this +on. Be aware that this will cause some of the tests run when you +type `make tests' to fail. + +There is a shell option, `xpg_echo', settable with `shopt', that will +change the behavior of echo at runtime. Enabling this option turns +on expansion of backslash-escape sequences. + +E6) Why doesn't a while or for loop get suspended when I type ^Z? + +This is a consequence of how job control works on Unix. The only +thing that can be suspended is the process group. This is a single +command or pipeline of commands that the shell forks and executes. + +When you run a while or for loop, the only thing that the shell forks +and executes are any commands in the while loop test and commands in +the loop bodies. These, therefore, are the only things that can be +suspended when you type ^Z. + +If you want to be able to stop the entire loop, you need to put it +within parentheses, which will force the loop into a subshell that +may be stopped (and subsequently restarted) as a single unit. + +E7) What about empty for loops in Makefiles? + +It's fairly common to see constructs like this in automatically-generated +Makefiles: + +SUBDIRS = @SUBDIRS@ + + ... + +subdirs-clean: + for d in ${SUBDIRS}; do \ + ( cd $$d && ${MAKE} ${MFLAGS} clean ) \ + done + +When SUBDIRS is empty, this results in a command like this being passed to +bash: + + for d in ; do + ( cd $d && ${MAKE} ${MFLAGS} clean ) + done + +In versions of bash before bash-2.05a, this was a syntax error. If the +reserved word `in' was present, a word must follow it before the semicolon +or newline. The language in the manual page referring to the list of words +being empty referred to the list after it is expanded. These versions of +bash required that there be at least one word following the `in' when the +construct was parsed. + +The idiomatic Makefile solution is something like: + +SUBDIRS = @SUBDIRS@ + +subdirs-clean: + subdirs=$SUBDIRS ; for d in $$subdirs; do \ + ( cd $$d && ${MAKE} ${MFLAGS} clean ) \ + done + +The latest drafts of the updated POSIX standard have changed this: the +word list is no longer required. Bash versions 2.05a and later accept +the new syntax. + +E8) Why does the arithmetic evaluation code complain about `08'? + +The bash arithmetic evaluation code (used for `let', $(()), (()), and in +other places), interprets a leading `0' in numeric constants as denoting +an octal number, and a leading `0x' as denoting hexadecimal. This is +in accordance with the POSIX.2 spec, section 2.9.2.1, which states that +arithmetic constants should be handled as signed long integers as defined +by the ANSI/ISO C standard. + +The POSIX.2 interpretation committee has confirmed this: + +http://www.pasc.org/interps/unofficial/db/p1003.2/pasc-1003.2-173.html + +E9) Why does the pattern matching expression [A-Z]* match files beginning + with every letter except `z'? + +Bash-2.03, Bash-2.05 and later versions honor the current locale setting +when processing ranges within pattern matching bracket expressions ([A-Z]). +This is what POSIX.2 and SUSv3/XPG6 specify. + +The behavior of the matcher in bash-2.05 and later versions depends on the +current LC_COLLATE setting. Setting this variable to `C' or `POSIX' will +result in the traditional behavior ([A-Z] matches all uppercase ASCII +characters). Many other locales, including the en_US locale (the default +on many US versions of Linux) collate the upper and lower case letters like +this: + + AaBb...Zz + +which means that [A-Z] matches every letter except `z'. Others collate like + + aAbBcC...zZ + +which means that [A-Z] matches every letter except `a'. + +The portable way to specify upper case letters is [:upper:] instead of +A-Z; lower case may be specified as [:lower:] instead of a-z. + +Look at the manual pages for setlocale(3), strcoll(3), and, if it is +present, locale(1). If you have locale(1), you can use it to find +your current locale information even if you do not have any of the +LC_ variables set. + +My advice is to put + + export LC_COLLATE=C + +into /etc/profile and inspect any shell scripts run from cron for +constructs like [A-Z]. This will prevent things like + + rm [A-Z]* + +from removing every file in the current directory except those beginning +with `z' and still allow individual users to change the collation order. +Users may put the above command into their own profiles as well, of course. + +E10) Why does `cd //' leave $PWD as `//'? + +POSIX.2, in its description of `cd', says that *three* or more leading +slashes may be replaced with a single slash when canonicalizing the +current working directory. + +This is, I presume, for historical compatibility. Certain versions of +Unix, and early network file systems, used paths of the form +//hostname/path to access `path' on server `hostname'. + +E11) If I resize my xterm while another program is running, why doesn't bash + notice the change? + +This is another issue that deals with job control. + +The kernel maintains a notion of a current terminal process group. Members +of this process group (processes whose process group ID is equal to the +current terminal process group ID) receive terminal-generated signals like +SIGWINCH. (For more details, see the JOB CONTROL section of the bash +man page.) + +If a terminal is resized, the kernel sends SIGWINCH to each member of +the terminal's current process group (the `foreground' process group). + +When bash is running with job control enabled, each pipeline (which may be +a single command) is run in its own process group, different from bash's +process group. This foreground process group receives the SIGWINCH; bash +does not. Bash has no way of knowing that the terminal has been resized. + +There is a `checkwinsize' option, settable with the `shopt' builtin, that +will cause bash to check the window size and adjust its idea of the +terminal's dimensions each time a process stops or exits and returns control +of the terminal to bash. Enable it with `shopt -s checkwinsize'. + +Section F: Things to watch out for on certain Unix versions + +F1) Why can't I use command line editing in my `cmdtool'? + +The problem is `cmdtool' and bash fighting over the input. When +scrolling is enabled in a cmdtool window, cmdtool puts the tty in +`raw mode' to permit command-line editing using the mouse for +applications that cannot do it themselves. As a result, bash and +cmdtool each try to read keyboard input immediately, with neither +getting enough of it to be useful. + +This mode also causes cmdtool to not implement many of the +terminal functions and control sequences appearing in the +`sun-cmd' termcap entry. For a more complete explanation, see +that file examples/suncmd.termcap in the bash distribution. + +`xterm' is a better choice, and gets along with bash much more +smoothly. + +If you must use cmdtool, you can use the termcap description in +examples/suncmd.termcap. Set the TERMCAP variable to the terminal +description contained in that file, i.e. + +TERMCAP='Mu|sun-cmd:am:bs:km:pt:li#34:co#80:cl=^L:ce=\E[K:cd=\E[J:rs=\E[s:' + +Then export TERMCAP and start a new cmdtool window from that shell. +The bash command-line editing should behave better in the new +cmdtool. If this works, you can put the assignment to TERMCAP +in your bashrc file. + +F2) I built bash on Solaris 2. Why do globbing expansions and filename + completion chop off the first few characters of each filename? + +This is the consequence of building bash on SunOS 5 and linking +with the libraries in /usr/ucblib, but using the definitions +and structures from files in /usr/include. + +The actual conflict is between the dirent structure in +/usr/include/dirent.h and the struct returned by the version of +`readdir' in libucb.a (a 4.3-BSD style `struct direct'). + +Make sure you've got /usr/ccs/bin ahead of /usr/ucb in your $PATH +when configuring and building bash. This will ensure that you +use /usr/ccs/bin/cc or acc instead of /usr/ucb/cc and that you +link with libc before libucb. + +If you have installed the Sun C compiler, you may also need to +put /usr/ccs/bin and /opt/SUNWspro/bin into your $PATH before +/usr/ucb. + +F3) Why does bash dump core after I interrupt username completion or + `~user' tilde expansion on a machine running NIS? + +This is a famous and long-standing bug in the SunOS YP (sorry, NIS) +client library, which is part of libc. + +The YP library code keeps static state -- a pointer into the data +returned from the server. When YP initializes itself (setpwent), +it looks at this pointer and calls free on it if it's non-null. +So far, so good. + +If one of the YP functions is interrupted during getpwent (the +exact function is interpretwithsave()), and returns NULL, the +pointer is freed without being reset to NULL, and the function +returns. The next time getpwent is called, it sees that this +pointer is non-null, calls free, and the bash free() blows up +because it's being asked to free freed memory. + +The traditional Unix mallocs allow memory to be freed multiple +times; that's probably why this has never been fixed. You can +run configure with the `--without-gnu-malloc' option to use +the C library malloc and avoid the problem. + +F4) I'm running SVR4.2. Why is the line erased every time I type `@'? + +The `@' character is the default `line kill' character in most +versions of System V, including SVR4.2. You can change this +character to whatever you want using `stty'. For example, to +change the line kill character to control-u, type + + stty kill ^U + +where the `^' and `U' can be two separate characters. + +F5) Why does bash report syntax errors when my C News scripts use a + redirection before a subshell command? + +The actual command in question is something like + + < file ( command ) + +According to the grammar given in the POSIX.2 standard, this construct +is, in fact, a syntax error. Redirections may only precede `simple +commands'. A subshell construct such as the above is one of the shell's +`compound commands'. A redirection may only follow a compound command. + +This affects the mechanical transformation of commands that use `cat' +to pipe a file into a command (a favorite Useless-Use-Of-Cat topic on +comp.unix.shell). While most commands of the form + + cat file | command + +can be converted to `< file command', shell control structures such as +loops and subshells require `command < file'. + +The file CWRU/sh-redir-hack in the bash-2.05a distribution is an +(unofficial) patch to parse.y that will modify the grammar to +support this construct. It will not apply with `patch'; you must +modify parse.y by hand. Note that if you apply this, you must +recompile with -DREDIRECTION_HACK. This introduces a large +number of reduce/reduce conflicts into the shell grammar. + +F6) Why can't I use vi-mode editing on Red Hat Linux 6.1? + +The short answer is that Red Hat screwed up. + +The long answer is that they shipped an /etc/inputrc that only works +for emacs mode editing, and then screwed all the vi users by setting +INPUTRC to /etc/inputrc in /etc/profile. + +The short fix is to do one of the following: remove or rename +/etc/inputrc, set INPUTRC=~/.inputrc in ~/.bashrc (or .bash_profile, +but make sure you export it if you do), remove the assignment to +INPUTRC from /etc/profile, add + + set keymap emacs + +to the beginning of /etc/inputrc, or bracket the key bindings in +/etc/inputrc with these lines + + $if mode=emacs + [...] + $endif + +F7) Why do bash-2.05a and bash-2.05b fail to compile `printf.def' on + HP/UX 11.x? + +HP/UX's support for long double is imperfect at best. + +GCC will support it without problems, but the HP C library functions +like strtold(3) and printf(3) don't actually work with long doubles. +HP implemented a `long_double' type as a 4-element array of 32-bit +ints, and that is what the library functions use. The ANSI C +`long double' type is a 128-bit floating point scalar. + +The easiest fix, until HP fixes things up, is to edit the generated +config.h and #undef the HAVE_LONG_DOUBLE line. After doing that, +the compilation should complete successfully. + +Section G: How can I get bash to do certain common things? + +G1) How can I get bash to read and display eight-bit characters? + +This is a process requiring several steps. + +First, you must ensure that the `physical' data path is a full eight +bits. For xterms, for example, the `vt100' resources `eightBitInput' +and `eightBitOutput' should be set to `true'. + +Once you have set up an eight-bit path, you must tell the kernel and +tty driver to leave the eighth bit of characters alone when processing +keyboard input. Use `stty' to do this: + + stty cs8 -istrip -parenb + +For old BSD-style systems, you can use + + stty pass8 + +You may also need + + stty even odd + +Finally, you need to tell readline that you will be inputting and +displaying eight-bit characters. You use readline variables to do +this. These variables can be set in your .inputrc or using the bash +`bind' builtin. Here's an example using `bind': + + bash$ bind 'set convert-meta off' + bash$ bind 'set meta-flag on' + bash$ bind 'set output-meta on' + +The `set' commands between the single quotes may also be placed +in ~/.inputrc. + +G2) How do I write a function `x' to replace builtin command `x', but + still invoke the command from within the function? + +This is why the `command' and `builtin' builtins exist. The +`command' builtin executes the command supplied as its first +argument, skipping over any function defined with that name. The +`builtin' builtin executes the builtin command given as its first +argument directly. + +For example, to write a function to replace `cd' that writes the +hostname and current directory to an xterm title bar, use +something like the following: + + cd() + { + builtin cd "$@" && xtitle "$HOST: $PWD" + } + +This could also be written using `command' instead of `builtin'; +the version above is marginally more efficient. + +G3) How can I find the value of a shell variable whose name is the value + of another shell variable? + +Versions of Bash newer than Bash-2.0 support this directly. You can use + + ${!var} + +For example, the following sequence of commands will echo `z': + + var1=var2 + var2=z + echo ${!var1} + +For sh compatibility, use the `eval' builtin. The important +thing to remember is that `eval' expands the arguments you give +it again, so you need to quote the parts of the arguments that +you want `eval' to act on. + +For example, this expression prints the value of the last positional +parameter: + + eval echo \"\$\{$#\}\" + +The expansion of the quoted portions of this expression will be +deferred until `eval' runs, while the `$#' will be expanded +before `eval' is executed. In versions of bash later than bash-2.0, + + echo ${!#} + +does the same thing. + +This is not the same thing as ksh93 `nameref' variables, though the syntax +is similar. I may add namerefs in a future bash version. + +G4) How can I make the bash `time' reserved word print timing output that + looks like the output from my system's /usr/bin/time? + +The bash command timing code looks for a variable `TIMEFORMAT' and +uses its value as a format string to decide how to display the +timing statistics. + +The value of TIMEFORMAT is a string with `%' escapes expanded in a +fashion similar in spirit to printf(3). The manual page explains +the meanings of the escape sequences in the format string. + +If TIMEFORMAT is not set, bash acts as if the following assignment had +been performed: + + TIMEFORMAT=$'\nreal\t%3lR\nuser\t%3lU\nsys\t%3lS' + +The POSIX.2 default time format (used by `time -p command') is + + TIMEFORMAT=$'real %2R\nuser %2U\nsys %2S' + +The BSD /usr/bin/time format can be emulated with: + + TIMEFORMAT=$'\t%1R real\t%1U user\t%1S sys' + +The System V /usr/bin/time format can be emulated with: + + TIMEFORMAT=$'\nreal\t%1R\nuser\t%1U\nsys\t%1S' + +The ksh format can be emulated with: + + TIMEFORMAT=$'\nreal\t%2lR\nuser\t%2lU\nsys\t%2lS' + +G5) How do I get the current directory into my prompt? + +Bash provides a number of backslash-escape sequences which are expanded +when the prompt string (PS1 or PS2) is displayed. The full list is in +the manual page. + +The \w expansion gives the full pathname of the current directory, with +a tilde (`~') substituted for the current value of $HOME. The \W +expansion gives the basename of the current directory. To put the full +pathname of the current directory into the path without any tilde +subsitution, use $PWD. Here are some examples: + + PS1='\w$ ' # current directory with tilde + PS1='\W$ ' # basename of current directory + PS1='$PWD$ ' # full pathname of current directory + +The single quotes are important in the final example to prevent $PWD from +being expanded when the assignment to PS1 is performed. + +G6) How can I rename "*.foo" to "*.bar"? + +Use the pattern removal functionality described in D3. The following `for' +loop will do the trick: + + for f in *.foo; do + mv $f ${f%foo}bar + done + +G7) How can I translate a filename from uppercase to lowercase? + +The script examples/functions/lowercase, originally written by John DuBois, +will do the trick. The converse is left as an exercise. + +G8) How can I write a filename expansion (globbing) pattern that will match + all files in the current directory except "." and ".."? + +You must have set the `extglob' shell option using `shopt -s extglob' to use +this: + + echo .!(.|) * + +A solution that works without extended globbing is given in the Unix Shell +FAQ, posted periodically to comp.unix.shell. + +Section H: Where do I go from here? + +H1) How do I report bugs in bash, and where should I look for fixes and + advice? + +Use the `bashbug' script to report bugs. It is built and +installed at the same time as bash. It provides a standard +template for reporting a problem and automatically includes +information about your configuration and build environment. + +`bashbug' sends its reports to bug-bash@gnu.org, which +is a large mailing list gatewayed to the usenet newsgroup gnu.bash.bug. + +Bug fixes, answers to questions, and announcements of new releases +are all posted to gnu.bash.bug. Discussions concerning bash features +and problems also take place there. + +To reach the bash maintainers directly, send mail to +bash-maintainers@gnu.org. + +H2) What kind of bash documentation is there? + +First, look in the doc directory in the bash distribution. It should +contain at least the following files: + +bash.1 an extensive, thorough Unix-style manual page +builtins.1 a manual page covering just bash builtin commands +bashref.texi a reference manual in GNU tex`info format +bashref.info an info version of the reference manual +FAQ this file +article.ms text of an article written for The Linux Journal +readline.3 a man page describing readline + +Postscript, HTML, and ASCII files created from the above source are +available in the documentation distribution. + +There is additional documentation available for anonymous FTP from host +ftp.cwru.edu in the `pub/bash' directory. + +Cameron Newham and Bill Rosenblatt have written a book on bash, published +by O'Reilly and Associates. The book is based on Bill Rosenblatt's Korn +Shell book. The title is ``Learning the Bash Shell'', and the ISBN number +is 1-56592-147-X. Look for it in fine bookstores near you. This book +covers bash-1.14, but has an appendix describing some of the new features +in bash-2.0. + +A second edition of this book is available, published in January, 1998. +The ISBN number is 1-56592-347-2. Look for it in the same fine bookstores +or on the web. + +The GNU Bash Reference Manual has been published as a printed book by +Network Theory Ltd (Paperback, ISBN: 0-9541617-7-7, Feb 2003). It covers +bash-2.0 and is available from most online bookstores (see +http://www.network-theory.co.uk/bash/manual/ for details). The publisher +will donate $1 to the Free Software Foundation for each copy sold. + +H3) What's coming in future versions? + +These are features I hope to include in a future version of bash. + +a better bash debugger (a minimally-tested version is included with bash-2.05b) +associative arrays +co-processes, but with a new-style syntax that looks like function declaration + +H4) What's on the bash `wish list' for future versions? + +These are features that may or may not appear in a future version of bash. + +breaking some of the shell functionality into embeddable libraries +a module system like zsh's, using dynamic loading like builtins +better internationalization using GNU `gettext' +date-stamped command history +a bash programmer's guide with a chapter on creating loadable builtins +a better loadable interface to perl with access to the shell builtins and + variables (contributions gratefully accepted) +ksh93-like `nameref' variables +ksh93-like `+=' variable assignment operator +ksh93-like `xx.yy' variables (including some of the .sh.* variables) and + associated disipline functions +Some of the new ksh93 pattern matching operators, like backreferencing + +H5) When will the next release appear? + +The next version will appear sometime in 2002. Never make predictions. + + +This document is Copyright 1995-2003 by Chester Ramey. + +Permission is hereby granted, without written agreement and +without license or royalty fees, to use, copy, and distribute +this document for any purpose, provided that the above copyright +notice appears in all copies of this document and that the +contents of this document remain unaltered. diff --git a/doc/aosa-bash.pdf.old b/doc/aosa-bash.pdf.old new file mode 100644 index 000000000..006a76776 Binary files /dev/null and b/doc/aosa-bash.pdf.old differ diff --git a/examples/loadables/Makefile.in.save b/examples/loadables/Makefile.in.save new file mode 100644 index 000000000..f6208f5cc --- /dev/null +++ b/examples/loadables/Makefile.in.save @@ -0,0 +1,238 @@ +# +# Simple makefile for the sample loadable builtins +# +# Copyright (C) 1996 Free Software Foundation, Inc. + +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation; either version 2, or (at your option) +# any later version. + +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. + +# You should have received a copy of the GNU General Public License +# along with this program; if not, write to the Free Software +# Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111 USA. + +# Include some boilerplate Gnu makefile definitions. +prefix = @prefix@ + +exec_prefix = @exec_prefix@ +bindir = @bindir@ +libdir = @libdir@ +infodir = @infodir@ +includedir = @includedir@ + +topdir = @top_srcdir@ +BUILD_DIR = @BUILD_DIR@ +srcdir = @srcdir@ +VPATH = .:@srcdir@ + +@SET_MAKE@ +CC = @CC@ +RM = rm -f + +SHELL = @MAKE_SHELL@ + +host_os = @host_os@ +host_cpu = @host_cpu@ +host_vendor = @host_vendor@ + +CFLAGS = @CFLAGS@ +LOCAL_CFLAGS = @LOCAL_CFLAGS@ +DEFS = @DEFS@ +LOCAL_DEFS = @LOCAL_DEFS@ + +CPPFLAGS = @CPPFLAGS@ + +BASHINCDIR = ${topdir}/include + +LIBBUILD = ${BUILD_DIR}/lib + +INTL_LIBSRC = ${topdir}/lib/intl +INTL_BUILDDIR = ${LIBBUILD}/intl +INTL_INC = @INTL_INC@ +LIBINTL_H = @LIBINTL_H@ + +CCFLAGS = $(DEFS) $(LOCAL_DEFS) $(LOCAL_CFLAGS) $(CFLAGS) + +# +# These values are generated for configure by ${topdir}/support/shobj-conf. +# If your system is not supported by that script, but includes facilities for +# dynamic loading of shared objects, please update the script and send the +# changes to bash-maintainers@gnu.org. +# +SHOBJ_CC = @SHOBJ_CC@ +SHOBJ_CFLAGS = @SHOBJ_CFLAGS@ +SHOBJ_LD = @SHOBJ_LD@ +SHOBJ_LDFLAGS = @SHOBJ_LDFLAGS@ +SHOBJ_XLDFLAGS = @SHOBJ_XLDFLAGS@ +SHOBJ_LIBS = @SHOBJ_LIBS@ +SHOBJ_STATUS = @SHOBJ_STATUS@ + +INC = -I. -I.. -I$(topdir) -I$(topdir)/lib -I$(topdir)/builtins \ + -I$(BASHINCDIR) -I$(BUILD_DIR) -I$(LIBBUILD) \ + -I$(BUILD_DIR)/builtins $(INTL_INC) + +.c.o: + $(SHOBJ_CC) $(SHOBJ_CFLAGS) $(CCFLAGS) $(INC) -c -o $@ $< + + +ALLPROG = print truefalse sleep pushd finfo logname basename dirname \ + tty pathchk tee head mkdir rmdir printenv id whoami \ + uname sync push ln unlink cut realpath getconf strftime +OTHERPROG = necho hello cat + +all: $(SHOBJ_STATUS) + +supported: $(ALLPROG) +others: $(OTHERPROG) + +unsupported: + @echo "Your system (${host_os}) is not supported by the" + @echo "${topdir}/support/shobj-conf script." + @echo "If your operating system provides facilities for dynamic" + @echo "loading of shared objects using the dlopen(3) interface," + @echo "please update the script and re-run configure. + @echo "Please send the changes you made to bash-maintainers@gnu.org" + @echo "for inclusion in future bash releases." + +everything: supported others + +print: print.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ print.o $(SHOBJ_LIBS) + +necho: necho.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ necho.o $(SHOBJ_LIBS) + +getconf: getconf.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ getconf.o $(SHOBJ_LIBS) + +hello: hello.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ hello.o $(SHOBJ_LIBS) + +truefalse: truefalse.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ truefalse.o $(SHOBJ_LIBS) + +sleep: sleep.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ sleep.o $(SHOBJ_LIBS) + +finfo: finfo.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ finfo.o $(SHOBJ_LIBS) + +cat: cat.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ cat.o $(SHOBJ_LIBS) + +logname: logname.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ logname.o $(SHOBJ_LIBS) + +basename: basename.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ basename.o $(SHOBJ_LIBS) + +dirname: dirname.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ dirname.o $(SHOBJ_LIBS) + +tty: tty.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ tty.o $(SHOBJ_LIBS) + +pathchk: pathchk.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ pathchk.o $(SHOBJ_LIBS) + +tee: tee.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ tee.o $(SHOBJ_LIBS) + +mkdir: mkdir.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ mkdir.o $(SHOBJ_LIBS) + +rmdir: rmdir.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ rmdir.o $(SHOBJ_LIBS) + +head: head.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ head.o $(SHOBJ_LIBS) + +printenv: printenv.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ printenv.o $(SHOBJ_LIBS) + +id: id.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ id.o $(SHOBJ_LIBS) + +whoami: whoami.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ whoami.o $(SHOBJ_LIBS) + +uname: uname.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ uname.o $(SHOBJ_LIBS) + +sync: sync.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ sync.o $(SHOBJ_LIBS) + +push: push.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ push.o $(SHOBJ_LIBS) + +ln: ln.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ ln.o $(SHOBJ_LIBS) + +unlink: unlink.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ unlink.o $(SHOBJ_LIBS) + +cut: cut.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ cut.o $(SHOBJ_LIBS) + +realpath: realpath.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ realpath.o $(SHOBJ_LIBS) + +strftime: strftime.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ strftime.o $(SHOBJ_LIBS) + +# pushd is a special case. We use the same source that the builtin version +# uses, with special compilation options. +# +pushd.c: ${topdir}/builtins/pushd.def + $(RM) $@ + ${BUILD_DIR}/builtins/mkbuiltins -D ${topdir}/builtins ${topdir}/builtins/pushd.def + +pushd.o: pushd.c + $(RM) $@ + $(SHOBJ_CC) -DHAVE_CONFIG_H -DPUSHD_AND_POPD -DLOADABLE_BUILTIN $(SHOBJ_CFLAGS) $(CFLAGS) $(CPPFLAGS) $(INC) -c -o $@ $< + +pushd: pushd.o + $(SHOBJ_LD) $(SHOBJ_LDFLAGS) $(SHOBJ_XLDFLAGS) -o $@ pushd.o $(SHOBJ_LIBS) + +clean: + $(RM) $(ALLPROG) $(OTHERPROG) *.o + -( cd perl && ${MAKE} ${MFLAGS} $@ ) + +mostlyclean: clean + -( cd perl && ${MAKE} ${MFLAGS} $@ ) + +distclean maintainer-clean: clean + $(RM) Makefile pushd.c + -( cd perl && ${MAKE} ${MFLAGS} $@ ) + +print.o: print.c +truefalse.o: truefalse.c +sleep.o: sleep.c +finfo.o: finfo.c +logname.o: logname.c +basename.o: basename.c +dirname.o: dirname.c +tty.o: tty.c +pathchk.o: pathchk.c +tee.o: tee.c +head.o: head.c +rmdir.o: rmdir.c +necho.o: necho.c +getconf.o: getconf.c +hello.o: hello.c +cat.o: cat.c +printenv.o: printenv.c +id.o: id.c +whoami.o: whoami.c +uname.o: uname.c +sync.o: sync.c +push.o: push.c +mkdir.o: mkdir.c +realpath.o: realpath.c +strftime.o: strftime.c diff --git a/lib/readline/doc/Makefile.old b/lib/readline/doc/Makefile.old new file mode 100644 index 000000000..58d4dd762 --- /dev/null +++ b/lib/readline/doc/Makefile.old @@ -0,0 +1,76 @@ +# This makefile for Readline library documentation is in -*- text -*- mode. +# Emacs likes it that way. +RM = rm -f + +MAKEINFO = makeinfo +TEXI2DVI = texi2dvi +TEXI2HTML = texi2html +QUIETPS = #set this to -q to shut up dvips +DVIPS = dvips -D 300 $(QUIETPS) -o $@ # tricky + +INSTALL_DATA = cp +infodir = /usr/local/info + +RLSRC = rlman.texinfo rluser.texinfo rltech.texinfo +HISTSRC = hist.texinfo hsuser.texinfo hstech.texinfo + +DVIOBJ = readline.dvi history.dvi +INFOOBJ = readline.info history.info +PSOBJ = readline.ps history.ps +HTMLOBJ = readline.html history.html + +all: info dvi html ps +nodvi: info html + +readline.dvi: $(RLSRC) + $(TEXI2DVI) rlman.texinfo + mv rlman.dvi readline.dvi + +readline.info: $(RLSRC) + $(MAKEINFO) --no-split -o $@ rlman.texinfo + +history.dvi: ${HISTSRC} + $(TEXI2DVI) hist.texinfo + mv hist.dvi history.dvi + +history.info: ${HISTSRC} + $(MAKEINFO) --no-split -o $@ hist.texinfo + +readline.ps: readline.dvi + $(RM) $@ + $(DVIPS) readline.dvi + +history.ps: history.dvi + $(RM) $@ + $(DVIPS) history.dvi + +readline.html: ${RLSRC} + $(TEXI2HTML) rlman.texinfo + sed -e 's:rlman.html:readline.html:' -e 's:rlman_toc.html:readline_toc.html:' rlman.html > readline.html + sed -e 's:rlman.html:readline.html:' -e 's:rlman_toc.html:readline_toc.html:' rlman_toc.html > readline_toc.html + $(RM) rlman.html rlman_toc.html + +history.html: ${HISTSRC} + $(TEXI2HTML) hist.texinfo + sed -e 's:hist.html:history.html:' -e 's:hist_toc.html:history_toc.html:' hist.html > history.html + sed -e 's:hist.html:history.html:' -e 's:hist_toc.html:history_toc.html:' hist_toc.html > history_toc.html + $(RM) hist.html hist_toc.html + +info: $(INFOOBJ) +dvi: $(DVIOBJ) +ps: $(PSOBJ) +html: $(HTMLOBJ) + +clean: + $(RM) *.aux *.cp *.fn *.ky *.log *.pg *.toc *.tp *.vr *.cps *.pgs \ + *.fns *.kys *.tps *.vrs *.o core + +distclean: clean +mostlyclean: clean + +maintainer-clean: clean + $(RM) *.dvi *.info *.info-* *.ps *.html + +install: info + ${INSTALL_DATA} readline.info $(infodir)/readline.info + ${INSTALL_DATA} history.info $(infodir)/history.info diff --git a/po/tr.po b/po/tr.po index 759d6f330..ec6f27424 100644 --- a/po/tr.po +++ b/po/tr.po @@ -9,7 +9,7 @@ msgstr "" "Project-Id-Version: bash 4.3-rc2\n" "Report-Msgid-Bugs-To: \n" "POT-Creation-Date: 2014-01-23 16:04-0500\n" -"PO-Revision-Date: 2014-03-22 14:56+0100\n" +"PO-Revision-Date: 2014-10-24 16:11+0200\n" "Last-Translator: Volkan Gezer \n" "Language-Team: Turkish \n" "Language: tr\n" @@ -4351,7 +4351,6 @@ msgstr "" " Son çalıştırılan komutun durumunu döndürür." #: builtins.c:1574 -#, fuzzy msgid "" "Report time consumed by pipeline's execution.\n" " \n" @@ -4366,13 +4365,18 @@ msgid "" " Exit Status:\n" " The return status is the return status of PIPELINE." msgstr "" -"\rtime [-p] [!] KOMUT1 [| KOMUT2 ...]\n" -" Bir boruhattındaki her komut bir kanal ile sonrakine bağlıdır. Böylece\n" -" her komut bir önceki komutun çıktısını okur. time boruhattındaki iş\n" -" sonuçlanınca geçen gerçek zaman, kullanıcı ve sistem zamanını basar.\n" -" Dönüş durumu boruhattının dönüş durumudur. -p seçeneği zaman\n" -" istatistiklerinin biraz farklı bir biçimde basılmasını sağlar; çıktı\n" -" biçimi olarak TIMEFORMAT değişkeninin değerini kullanır." +"Veriyolu çalıştırması tarafından harcanan zamanı bildir.\n" +" \n" +" VERİYOLUnu çalıştır ve sonlandığında kullanıcı işlemci zamanı ile sistem\n" +" işlemci zamanını içeren gerçek zamanın bir özetini yazdır.\n" +" \n" +" Seçenekler:\n" +" -p\ttaşınabilir Posix biçiminde zaman özetini yaz\n" +" \n" +" ZAMANBİÇİMİ değişkeninin değeri çıktı biçimi olarak kullanılır.\n" +" \n" +" Çıktı Durumu:\n" +" VERİYOLU döndürme değerini döndürür." #: builtins.c:1591 msgid "" @@ -4423,7 +4427,6 @@ msgstr "" " doğru sonuç vermemişse sıfır döner." #: builtins.c:1620 -#, fuzzy msgid "" "Execute commands as long as a test succeeds.\n" " \n" @@ -4433,9 +4436,13 @@ msgid "" " Exit Status:\n" " Returns the status of the last command executed." msgstr "" -"\rwhile KOMUTlar; do KOMUTlar; done\n" -" `while KOMUTlar; listesinin çıkış durumu sıfır olduğu sürece\n" -" `do KOMUTlar;' listesi çalıştırılır." +"Testler başarılı olduğu sürece komutları çalıştırır.\n" +" \n" +" KOMUTLARı, `while' KOMUTLARı sıfır çıktı durumuna sahip oluncaya\n" +" kadar genişlet ve çalıştır.\n" +" \n" +" Çıktı Durumu:\n" +" En son çalıştırılan komutun durumunu döndürür." #: builtins.c:1632 msgid "" diff --git a/subst.c b/subst.c index d89660734..1601f4df3 100644 --- a/subst.c +++ b/subst.c @@ -3494,9 +3494,12 @@ expand_string_for_rhs (string, quoted, dollar_at_p, has_dollar_at) if (string == 0 || *string == '\0') return (WORD_LIST *)NULL; + expand_no_split_dollar_star = 1; td.flags = W_NOSPLIT2; /* no splitting, remove "" and '' */ td.word = string; tresult = call_expand_word_internal (&td, quoted, 1, dollar_at_p, has_dollar_at); + expand_no_split_dollar_star = 0; + return (tresult); } @@ -6045,6 +6048,13 @@ parameter_brace_expand_rhs (name, value, c, quoted, qdollaratp, hasdollarat) } else if ((quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) && hasdol) { + /* Posix interp 221 changed the rules on this. The idea is that + something like "$xxx$@" should expand the same as "${foo-$xxx$@}" + when foo and xxx are unset. The problem is that it's not in any + way backwards compatible and few other shells do it. We're eventually + going to try and split the difference (heh) a little bit here. */ + /* hasdol == 1 means we saw a quoted dollar at. */ + /* The brace expansion occurred between double quotes and there was a $@ in TEMP. It does not matter if the $@ is quoted, as long as it does not expand to anything. In this case, we want to return @@ -6094,10 +6104,8 @@ parameter_brace_expand_rhs (name, value, c, quoted, qdollaratp, hasdollarat) else #endif /* ARRAY_VARS */ bind_variable (vname, t1, 0); -#if 0 - if (STREQ (vname, "IFS") == 0) -#endif - stupidly_hack_special_variables (vname); + + stupidly_hack_special_variables (vname); if (vname != name) free (vname); @@ -7862,10 +7870,11 @@ param_expand (string, sindex, quoted, expanded_something, temp = (quoted & (Q_DOUBLE_QUOTES|Q_PATQUOTE)) ? string_list_dollar_star (list) : string_list (list); if (temp) { - temp1 = quote_string (temp); + temp1 = (quoted & Q_DOUBLE_QUOTES) ? quote_string (temp) : temp; if (*temp == 0) tflag |= W_HASQUOTEDNULL; - free (temp); + if (temp != temp1) + free (temp); temp = temp1; } } diff --git a/subst.c~ b/subst.c~ new file mode 100644 index 000000000..8084f0942 --- /dev/null +++ b/subst.c~ @@ -0,0 +1,9852 @@ +/* subst.c -- The part of the shell that does parameter, command, arithmetic, + and globbing substitutions. */ + +/* ``Have a little faith, there's magic in the night. You ain't a + beauty, but, hey, you're alright.'' */ + +/* Copyright (C) 1987-2013 Free Software Foundation, Inc. + + This file is part of GNU Bash, the Bourne Again SHell. + + Bash is free software: you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation, either version 3 of the License, or + (at your option) any later version. + + Bash is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with Bash. If not, see . +*/ + +#include "config.h" + +#include "bashtypes.h" +#include +#include "chartypes.h" +#if defined (HAVE_PWD_H) +# include +#endif +#include +#include + +#if defined (HAVE_UNISTD_H) +# include +#endif + +#include "bashansi.h" +#include "posixstat.h" +#include "bashintl.h" + +#include "shell.h" +#include "parser.h" +#include "flags.h" +#include "jobs.h" +#include "execute_cmd.h" +#include "filecntl.h" +#include "trap.h" +#include "pathexp.h" +#include "mailcheck.h" + +#include "shmbutil.h" +#include "typemax.h" + +#include "builtins/getopt.h" +#include "builtins/common.h" + +#include "builtins/builtext.h" + +#include +#include + +#if !defined (errno) +extern int errno; +#endif /* !errno */ + +/* The size that strings change by. */ +#define DEFAULT_INITIAL_ARRAY_SIZE 112 +#define DEFAULT_ARRAY_SIZE 128 + +/* Variable types. */ +#define VT_VARIABLE 0 +#define VT_POSPARMS 1 +#define VT_ARRAYVAR 2 +#define VT_ARRAYMEMBER 3 +#define VT_ASSOCVAR 4 + +#define VT_STARSUB 128 /* $* or ${array[*]} -- used to split */ + +/* Flags for quoted_strchr */ +#define ST_BACKSL 0x01 +#define ST_CTLESC 0x02 +#define ST_SQUOTE 0x04 /* unused yet */ +#define ST_DQUOTE 0x08 /* unused yet */ + +/* Flags for the `pflags' argument to param_expand() */ +#define PF_NOCOMSUB 0x01 /* Do not perform command substitution */ +#define PF_IGNUNBOUND 0x02 /* ignore unbound vars even if -u set */ +#define PF_NOSPLIT2 0x04 /* same as W_NOSPLIT2 */ +#define PF_ASSIGNRHS 0x08 /* same as W_ASSIGNRHS */ + +/* These defs make it easier to use the editor. */ +#define LBRACE '{' +#define RBRACE '}' +#define LPAREN '(' +#define RPAREN ')' +#define LBRACK '[' +#define RBRACK ']' + +#if defined (HANDLE_MULTIBYTE) +#define WLPAREN L'(' +#define WRPAREN L')' +#endif + +/* Evaluates to 1 if C is one of the shell's special parameters whose length + can be taken, but is also one of the special expansion characters. */ +#define VALID_SPECIAL_LENGTH_PARAM(c) \ + ((c) == '-' || (c) == '?' || (c) == '#') + +/* Evaluates to 1 if C is one of the shell's special parameters for which an + indirect variable reference may be made. */ +#define VALID_INDIR_PARAM(c) \ + ((posixly_correct == 0 && (c) == '#') || (posixly_correct == 0 && (c) == '?') || (c) == '@' || (c) == '*') + +/* Evaluates to 1 if C is one of the OP characters that follows the parameter + in ${parameter[:]OPword}. */ +#define VALID_PARAM_EXPAND_CHAR(c) (sh_syntaxtab[(unsigned char)c] & CSUBSTOP) + +/* Evaluates to 1 if this is one of the shell's special variables. */ +#define SPECIAL_VAR(name, wi) \ + ((DIGIT (*name) && all_digits (name)) || \ + (name[1] == '\0' && (sh_syntaxtab[(unsigned char)*name] & CSPECVAR)) || \ + (wi && name[2] == '\0' && VALID_INDIR_PARAM (name[1]))) + +/* An expansion function that takes a string and a quoted flag and returns + a WORD_LIST *. Used as the type of the third argument to + expand_string_if_necessary(). */ +typedef WORD_LIST *EXPFUNC __P((char *, int)); + +/* Process ID of the last command executed within command substitution. */ +pid_t last_command_subst_pid = NO_PID; +pid_t current_command_subst_pid = NO_PID; + +/* Variables used to keep track of the characters in IFS. */ +SHELL_VAR *ifs_var; +char *ifs_value; +unsigned char ifs_cmap[UCHAR_MAX + 1]; +int ifs_is_set, ifs_is_null; + +#if defined (HANDLE_MULTIBYTE) +unsigned char ifs_firstc[MB_LEN_MAX]; +size_t ifs_firstc_len; +#else +unsigned char ifs_firstc; +#endif + +/* Sentinel to tell when we are performing variable assignments preceding a + command name and putting them into the environment. Used to make sure + we use the temporary environment when looking up variable values. */ +int assigning_in_environment; + +/* Used to hold a list of variable assignments preceding a command. Global + so the SIGCHLD handler in jobs.c can unwind-protect it when it runs a + SIGCHLD trap and so it can be saved and restored by the trap handlers. */ +WORD_LIST *subst_assign_varlist = (WORD_LIST *)NULL; + +/* Extern functions and variables from different files. */ +extern int last_command_exit_value, last_command_exit_signal; +extern int subshell_environment, line_number; +extern int subshell_level, parse_and_execute_level, sourcelevel; +extern int eof_encountered; +extern int return_catch_flag, return_catch_value; +extern pid_t dollar_dollar_pid; +extern int posixly_correct; +extern char *this_command_name; +extern struct fd_bitmap *current_fds_to_close; +extern int wordexp_only; +extern int expanding_redir; +extern int tempenv_assign_error; +extern int builtin_ignoring_errexit; + +#if !defined (HAVE_WCSDUP) && defined (HANDLE_MULTIBYTE) +extern wchar_t *wcsdup __P((const wchar_t *)); +#endif + +/* Non-zero means to allow unmatched globbed filenames to expand to + a null file. */ +int allow_null_glob_expansion; + +/* Non-zero means to throw an error when globbing fails to match anything. */ +int fail_glob_expansion; + +#if 0 +/* Variables to keep track of which words in an expanded word list (the + output of expand_word_list_internal) are the result of globbing + expansions. GLOB_ARGV_FLAGS is used by execute_cmd.c. + (CURRENTLY UNUSED). */ +char *glob_argv_flags; +static int glob_argv_flags_size; +#endif + +static WORD_LIST expand_word_error, expand_word_fatal; +static WORD_DESC expand_wdesc_error, expand_wdesc_fatal; +static char expand_param_error, expand_param_fatal; +static char extract_string_error, extract_string_fatal; + +/* Tell the expansion functions to not longjmp back to top_level on fatal + errors. Enabled when doing completion and prompt string expansion. */ +static int no_longjmp_on_fatal_error = 0; + +/* Set by expand_word_unsplit; used to inhibit splitting and re-joining + $* on $IFS, primarily when doing assignment statements. */ +static int expand_no_split_dollar_star = 0; + +/* A WORD_LIST of words to be expanded by expand_word_list_internal, + without any leading variable assignments. */ +static WORD_LIST *garglist = (WORD_LIST *)NULL; + +static char *quoted_substring __P((char *, int, int)); +static int quoted_strlen __P((char *)); +static char *quoted_strchr __P((char *, int, int)); + +static char *expand_string_if_necessary __P((char *, int, EXPFUNC *)); +static inline char *expand_string_to_string_internal __P((char *, int, EXPFUNC *)); +static WORD_LIST *call_expand_word_internal __P((WORD_DESC *, int, int, int *, int *)); +static WORD_LIST *expand_string_internal __P((char *, int)); +static WORD_LIST *expand_string_leave_quoted __P((char *, int)); +static WORD_LIST *expand_string_for_rhs __P((char *, int, int *, int *)); + +static WORD_LIST *list_quote_escapes __P((WORD_LIST *)); +static WORD_LIST *list_dequote_escapes __P((WORD_LIST *)); + +static char *make_quoted_char __P((int)); +static WORD_LIST *quote_list __P((WORD_LIST *)); + +static int unquoted_substring __P((char *, char *)); +static int unquoted_member __P((int, char *)); + +#if defined (ARRAY_VARS) +static SHELL_VAR *do_compound_assignment __P((char *, char *, int)); +#endif +static int do_assignment_internal __P((const WORD_DESC *, int)); + +static char *string_extract_verbatim __P((char *, size_t, int *, char *, int)); +static char *string_extract __P((char *, int *, char *, int)); +static char *string_extract_double_quoted __P((char *, int *, int)); +static inline char *string_extract_single_quoted __P((char *, int *)); +static inline int skip_single_quoted __P((const char *, size_t, int)); +static int skip_double_quoted __P((char *, size_t, int)); +static char *extract_delimited_string __P((char *, int *, char *, char *, char *, int)); +static char *extract_dollar_brace_string __P((char *, int *, int, int)); +static int skip_matched_pair __P((const char *, int, int, int, int)); + +static char *pos_params __P((char *, int, int, int)); + +static unsigned char *mb_getcharlens __P((char *, int)); + +static char *remove_upattern __P((char *, char *, int)); +#if defined (HANDLE_MULTIBYTE) +static wchar_t *remove_wpattern __P((wchar_t *, size_t, wchar_t *, int)); +#endif +static char *remove_pattern __P((char *, char *, int)); + +static int match_upattern __P((char *, char *, int, char **, char **)); +#if defined (HANDLE_MULTIBYTE) +static int match_wpattern __P((wchar_t *, char **, size_t, wchar_t *, int, char **, char **)); +#endif +static int match_pattern __P((char *, char *, int, char **, char **)); +static int getpatspec __P((int, char *)); +static char *getpattern __P((char *, int, int)); +static char *variable_remove_pattern __P((char *, char *, int, int)); +static char *list_remove_pattern __P((WORD_LIST *, char *, int, int, int)); +static char *parameter_list_remove_pattern __P((int, char *, int, int)); +#ifdef ARRAY_VARS +static char *array_remove_pattern __P((SHELL_VAR *, char *, int, char *, int)); +#endif +static char *parameter_brace_remove_pattern __P((char *, char *, int, char *, int, int, int)); + +static char *process_substitute __P((char *, int)); + +static char *read_comsub __P((int, int, int *)); + +#ifdef ARRAY_VARS +static arrayind_t array_length_reference __P((char *)); +#endif + +static int valid_brace_expansion_word __P((char *, int)); +static int chk_atstar __P((char *, int, int *, int *)); +static int chk_arithsub __P((const char *, int)); + +static WORD_DESC *parameter_brace_expand_word __P((char *, int, int, int, arrayind_t *)); +static char *parameter_brace_find_indir __P((char *, int, int, int)); +static WORD_DESC *parameter_brace_expand_indir __P((char *, int, int, int *, int *)); +static WORD_DESC *parameter_brace_expand_rhs __P((char *, char *, int, int, int *, int *)); +static void parameter_brace_expand_error __P((char *, char *)); + +static int valid_length_expression __P((char *)); +static intmax_t parameter_brace_expand_length __P((char *)); + +static char *skiparith __P((char *, int)); +static int verify_substring_values __P((SHELL_VAR *, char *, char *, int, intmax_t *, intmax_t *)); +static int get_var_and_type __P((char *, char *, arrayind_t, int, int, SHELL_VAR **, char **)); +static char *mb_substring __P((char *, int, int)); +static char *parameter_brace_substring __P((char *, char *, int, char *, int, int)); + +static int shouldexp_replacement __P((char *)); + +static char *pos_params_pat_subst __P((char *, char *, char *, int)); + +static char *parameter_brace_patsub __P((char *, char *, int, char *, int, int)); + +static char *pos_params_casemod __P((char *, char *, int, int)); +static char *parameter_brace_casemod __P((char *, char *, int, int, char *, int, int)); + +static WORD_DESC *parameter_brace_expand __P((char *, int *, int, int, int *, int *)); +static WORD_DESC *param_expand __P((char *, int *, int, int *, int *, int *, int *, int)); + +static WORD_LIST *expand_word_internal __P((WORD_DESC *, int, int, int *, int *)); + +static WORD_LIST *word_list_split __P((WORD_LIST *)); + +static void exp_jump_to_top_level __P((int)); + +static WORD_LIST *separate_out_assignments __P((WORD_LIST *)); +static WORD_LIST *glob_expand_word_list __P((WORD_LIST *, int)); +#ifdef BRACE_EXPANSION +static WORD_LIST *brace_expand_word_list __P((WORD_LIST *, int)); +#endif +#if defined (ARRAY_VARS) +static int make_internal_declare __P((char *, char *)); +#endif +static WORD_LIST *shell_expand_word_list __P((WORD_LIST *, int)); +static WORD_LIST *expand_word_list_internal __P((WORD_LIST *, int)); + +/* **************************************************************** */ +/* */ +/* Utility Functions */ +/* */ +/* **************************************************************** */ + +#if defined (DEBUG) +void +dump_word_flags (flags) + int flags; +{ + int f; + + f = flags; + fprintf (stderr, "%d -> ", f); + if (f & W_ARRAYIND) + { + f &= ~W_ARRAYIND; + fprintf (stderr, "W_ARRAYIND%s", f ? "|" : ""); + } + if (f & W_ASSIGNASSOC) + { + f &= ~W_ASSIGNASSOC; + fprintf (stderr, "W_ASSIGNASSOC%s", f ? "|" : ""); + } + if (f & W_ASSIGNARRAY) + { + f &= ~W_ASSIGNARRAY; + fprintf (stderr, "W_ASSIGNARRAY%s", f ? "|" : ""); + } + if (f & W_HASCTLESC) + { + f &= ~W_HASCTLESC; + fprintf (stderr, "W_HASCTLESC%s", f ? "|" : ""); + } + if (f & W_NOPROCSUB) + { + f &= ~W_NOPROCSUB; + fprintf (stderr, "W_NOPROCSUB%s", f ? "|" : ""); + } + if (f & W_DQUOTE) + { + f &= ~W_DQUOTE; + fprintf (stderr, "W_DQUOTE%s", f ? "|" : ""); + } + if (f & W_HASQUOTEDNULL) + { + f &= ~W_HASQUOTEDNULL; + fprintf (stderr, "W_HASQUOTEDNULL%s", f ? "|" : ""); + } + if (f & W_ASSIGNARG) + { + f &= ~W_ASSIGNARG; + fprintf (stderr, "W_ASSIGNARG%s", f ? "|" : ""); + } + if (f & W_ASSNBLTIN) + { + f &= ~W_ASSNBLTIN; + fprintf (stderr, "W_ASSNBLTIN%s", f ? "|" : ""); + } + if (f & W_ASSNGLOBAL) + { + f &= ~W_ASSNGLOBAL; + fprintf (stderr, "W_ASSNGLOBAL%s", f ? "|" : ""); + } + if (f & W_ASSIGNINT) + { + f &= ~W_ASSIGNINT; + fprintf (stderr, "W_ASSIGNINT%s", f ? "|" : ""); + } + if (f & W_COMPASSIGN) + { + f &= ~W_COMPASSIGN; + fprintf (stderr, "W_COMPASSIGN%s", f ? "|" : ""); + } + if (f & W_NOEXPAND) + { + f &= ~W_NOEXPAND; + fprintf (stderr, "W_NOEXPAND%s", f ? "|" : ""); + } + if (f & W_ITILDE) + { + f &= ~W_ITILDE; + fprintf (stderr, "W_ITILDE%s", f ? "|" : ""); + } + if (f & W_NOTILDE) + { + f &= ~W_NOTILDE; + fprintf (stderr, "W_NOTILDE%s", f ? "|" : ""); + } + if (f & W_ASSIGNRHS) + { + f &= ~W_ASSIGNRHS; + fprintf (stderr, "W_ASSIGNRHS%s", f ? "|" : ""); + } + if (f & W_NOCOMSUB) + { + f &= ~W_NOCOMSUB; + fprintf (stderr, "W_NOCOMSUB%s", f ? "|" : ""); + } + if (f & W_DOLLARSTAR) + { + f &= ~W_DOLLARSTAR; + fprintf (stderr, "W_DOLLARSTAR%s", f ? "|" : ""); + } + if (f & W_DOLLARAT) + { + f &= ~W_DOLLARAT; + fprintf (stderr, "W_DOLLARAT%s", f ? "|" : ""); + } + if (f & W_TILDEEXP) + { + f &= ~W_TILDEEXP; + fprintf (stderr, "W_TILDEEXP%s", f ? "|" : ""); + } + if (f & W_NOSPLIT2) + { + f &= ~W_NOSPLIT2; + fprintf (stderr, "W_NOSPLIT2%s", f ? "|" : ""); + } + if (f & W_NOSPLIT) + { + f &= ~W_NOSPLIT; + fprintf (stderr, "W_NOSPLIT%s", f ? "|" : ""); + } + if (f & W_NOBRACE) + { + f &= ~W_NOBRACE; + fprintf (stderr, "W_NOBRACE%s", f ? "|" : ""); + } + if (f & W_NOGLOB) + { + f &= ~W_NOGLOB; + fprintf (stderr, "W_NOGLOB%s", f ? "|" : ""); + } + if (f & W_SPLITSPACE) + { + f &= ~W_SPLITSPACE; + fprintf (stderr, "W_SPLITSPACE%s", f ? "|" : ""); + } + if (f & W_ASSIGNMENT) + { + f &= ~W_ASSIGNMENT; + fprintf (stderr, "W_ASSIGNMENT%s", f ? "|" : ""); + } + if (f & W_QUOTED) + { + f &= ~W_QUOTED; + fprintf (stderr, "W_QUOTED%s", f ? "|" : ""); + } + if (f & W_HASDOLLAR) + { + f &= ~W_HASDOLLAR; + fprintf (stderr, "W_HASDOLLAR%s", f ? "|" : ""); + } + fprintf (stderr, "\n"); + fflush (stderr); +} +#endif + +#ifdef INCLUDE_UNUSED +static char * +quoted_substring (string, start, end) + char *string; + int start, end; +{ + register int len, l; + register char *result, *s, *r; + + len = end - start; + + /* Move to string[start], skipping quoted characters. */ + for (s = string, l = 0; *s && l < start; ) + { + if (*s == CTLESC) + { + s++; + continue; + } + l++; + if (*s == 0) + break; + } + + r = result = (char *)xmalloc (2*len + 1); /* save room for quotes */ + + /* Copy LEN characters, including quote characters. */ + s = string + l; + for (l = 0; l < len; s++) + { + if (*s == CTLESC) + *r++ = *s++; + *r++ = *s; + l++; + if (*s == 0) + break; + } + *r = '\0'; + return result; +} +#endif + +#ifdef INCLUDE_UNUSED +/* Return the length of S, skipping over quoted characters */ +static int +quoted_strlen (s) + char *s; +{ + register char *p; + int i; + + i = 0; + for (p = s; *p; p++) + { + if (*p == CTLESC) + { + p++; + if (*p == 0) + return (i + 1); + } + i++; + } + + return i; +} +#endif + +/* Find the first occurrence of character C in string S, obeying shell + quoting rules. If (FLAGS & ST_BACKSL) is non-zero, backslash-escaped + characters are skipped. If (FLAGS & ST_CTLESC) is non-zero, characters + escaped with CTLESC are skipped. */ +static char * +quoted_strchr (s, c, flags) + char *s; + int c, flags; +{ + register char *p; + + for (p = s; *p; p++) + { + if (((flags & ST_BACKSL) && *p == '\\') + || ((flags & ST_CTLESC) && *p == CTLESC)) + { + p++; + if (*p == '\0') + return ((char *)NULL); + continue; + } + else if (*p == c) + return p; + } + return ((char *)NULL); +} + +/* Return 1 if CHARACTER appears in an unquoted portion of + STRING. Return 0 otherwise. CHARACTER must be a single-byte character. */ +static int +unquoted_member (character, string) + int character; + char *string; +{ + size_t slen; + int sindex, c; + DECLARE_MBSTATE; + + slen = strlen (string); + sindex = 0; + while (c = string[sindex]) + { + if (c == character) + return (1); + + switch (c) + { + default: + ADVANCE_CHAR (string, slen, sindex); + break; + + case '\\': + sindex++; + if (string[sindex]) + ADVANCE_CHAR (string, slen, sindex); + break; + + case '\'': + sindex = skip_single_quoted (string, slen, ++sindex); + break; + + case '"': + sindex = skip_double_quoted (string, slen, ++sindex); + break; + } + } + return (0); +} + +/* Return 1 if SUBSTR appears in an unquoted portion of STRING. */ +static int +unquoted_substring (substr, string) + char *substr, *string; +{ + size_t slen; + int sindex, c, sublen; + DECLARE_MBSTATE; + + if (substr == 0 || *substr == '\0') + return (0); + + slen = strlen (string); + sublen = strlen (substr); + for (sindex = 0; c = string[sindex]; ) + { + if (STREQN (string + sindex, substr, sublen)) + return (1); + + switch (c) + { + case '\\': + sindex++; + if (string[sindex]) + ADVANCE_CHAR (string, slen, sindex); + break; + + case '\'': + sindex = skip_single_quoted (string, slen, ++sindex); + break; + + case '"': + sindex = skip_double_quoted (string, slen, ++sindex); + break; + + default: + ADVANCE_CHAR (string, slen, sindex); + break; + } + } + return (0); +} + +/* Most of the substitutions must be done in parallel. In order + to avoid using tons of unclear goto's, I have some functions + for manipulating malloc'ed strings. They all take INDX, a + pointer to an integer which is the offset into the string + where manipulation is taking place. They also take SIZE, a + pointer to an integer which is the current length of the + character array for this string. */ + +/* Append SOURCE to TARGET at INDEX. SIZE is the current amount + of space allocated to TARGET. SOURCE can be NULL, in which + case nothing happens. Gets rid of SOURCE by freeing it. + Returns TARGET in case the location has changed. */ +INLINE char * +sub_append_string (source, target, indx, size) + char *source, *target; + int *indx, *size; +{ + if (source) + { + int srclen, n; + + srclen = STRLEN (source); + if (srclen >= (int)(*size - *indx)) + { + n = srclen + *indx; + n = (n + DEFAULT_ARRAY_SIZE) - (n % DEFAULT_ARRAY_SIZE); + target = (char *)xrealloc (target, (*size = n)); + } + + FASTCOPY (source, target + *indx, srclen); + *indx += srclen; + target[*indx] = '\0'; + + free (source); + } + return (target); +} + +#if 0 +/* UNUSED */ +/* Append the textual representation of NUMBER to TARGET. + INDX and SIZE are as in SUB_APPEND_STRING. */ +char * +sub_append_number (number, target, indx, size) + intmax_t number; + int *indx, *size; + char *target; +{ + char *temp; + + temp = itos (number); + return (sub_append_string (temp, target, indx, size)); +} +#endif + +/* Extract a substring from STRING, starting at SINDEX and ending with + one of the characters in CHARLIST. Don't make the ending character + part of the string. Leave SINDEX pointing at the ending character. + Understand about backslashes in the string. If (flags & SX_VARNAME) + is non-zero, and array variables have been compiled into the shell, + everything between a `[' and a corresponding `]' is skipped over. + If (flags & SX_NOALLOC) is non-zero, don't return the substring, just + update SINDEX. If (flags & SX_REQMATCH) is non-zero, the string must + contain a closing character from CHARLIST. */ +static char * +string_extract (string, sindex, charlist, flags) + char *string; + int *sindex; + char *charlist; + int flags; +{ + register int c, i; + int found; + size_t slen; + char *temp; + DECLARE_MBSTATE; + + slen = (MB_CUR_MAX > 1) ? strlen (string + *sindex) + *sindex : 0; + i = *sindex; + found = 0; + while (c = string[i]) + { + if (c == '\\') + { + if (string[i + 1]) + i++; + else + break; + } +#if defined (ARRAY_VARS) + else if ((flags & SX_VARNAME) && c == '[') + { + int ni; + /* If this is an array subscript, skip over it and continue. */ + ni = skipsubscript (string, i, 0); + if (string[ni] == ']') + i = ni; + } +#endif + else if (MEMBER (c, charlist)) + { + found = 1; + break; + } + + ADVANCE_CHAR (string, slen, i); + } + + /* If we had to have a matching delimiter and didn't find one, return an + error and let the caller deal with it. */ + if ((flags & SX_REQMATCH) && found == 0) + { + *sindex = i; + return (&extract_string_error); + } + + temp = (flags & SX_NOALLOC) ? (char *)NULL : substring (string, *sindex, i); + *sindex = i; + + return (temp); +} + +/* Extract the contents of STRING as if it is enclosed in double quotes. + SINDEX, when passed in, is the offset of the character immediately + following the opening double quote; on exit, SINDEX is left pointing after + the closing double quote. If STRIPDQ is non-zero, unquoted double + quotes are stripped and the string is terminated by a null byte. + Backslashes between the embedded double quotes are processed. If STRIPDQ + is zero, an unquoted `"' terminates the string. */ +static char * +string_extract_double_quoted (string, sindex, stripdq) + char *string; + int *sindex, stripdq; +{ + size_t slen; + char *send; + int j, i, t; + unsigned char c; + char *temp, *ret; /* The new string we return. */ + int pass_next, backquote, si; /* State variables for the machine. */ + int dquote; + DECLARE_MBSTATE; + + slen = strlen (string + *sindex) + *sindex; + send = string + slen; + + pass_next = backquote = dquote = 0; + temp = (char *)xmalloc (1 + slen - *sindex); + + j = 0; + i = *sindex; + while (c = string[i]) + { + /* Process a character that was quoted by a backslash. */ + if (pass_next) + { + /* XXX - take another look at this in light of Interp 221 */ + /* Posix.2 sez: + + ``The backslash shall retain its special meaning as an escape + character only when followed by one of the characters: + $ ` " \ ''. + + If STRIPDQ is zero, we handle the double quotes here and let + expand_word_internal handle the rest. If STRIPDQ is non-zero, + we have already been through one round of backslash stripping, + and want to strip these backslashes only if DQUOTE is non-zero, + indicating that we are inside an embedded double-quoted string. */ + + /* If we are in an embedded quoted string, then don't strip + backslashes before characters for which the backslash + retains its special meaning, but remove backslashes in + front of other characters. If we are not in an + embedded quoted string, don't strip backslashes at all. + This mess is necessary because the string was already + surrounded by double quotes (and sh has some really weird + quoting rules). + The returned string will be run through expansion as if + it were double-quoted. */ + if ((stripdq == 0 && c != '"') || + (stripdq && ((dquote && (sh_syntaxtab[c] & CBSDQUOTE)) || dquote == 0))) + temp[j++] = '\\'; + pass_next = 0; + +add_one_character: + COPY_CHAR_I (temp, j, string, send, i); + continue; + } + + /* A backslash protects the next character. The code just above + handles preserving the backslash in front of any character but + a double quote. */ + if (c == '\\') + { + pass_next++; + i++; + continue; + } + + /* Inside backquotes, ``the portion of the quoted string from the + initial backquote and the characters up to the next backquote + that is not preceded by a backslash, having escape characters + removed, defines that command''. */ + if (backquote) + { + if (c == '`') + backquote = 0; + temp[j++] = c; + i++; + continue; + } + + if (c == '`') + { + temp[j++] = c; + backquote++; + i++; + continue; + } + + /* Pass everything between `$(' and the matching `)' or a quoted + ${ ... } pair through according to the Posix.2 specification. */ + if (c == '$' && ((string[i + 1] == LPAREN) || (string[i + 1] == LBRACE))) + { + int free_ret = 1; + + si = i + 2; + if (string[i + 1] == LPAREN) + ret = extract_command_subst (string, &si, 0); + else + ret = extract_dollar_brace_string (string, &si, Q_DOUBLE_QUOTES, 0); + + temp[j++] = '$'; + temp[j++] = string[i + 1]; + + /* Just paranoia; ret will not be 0 unless no_longjmp_on_fatal_error + is set. */ + if (ret == 0 && no_longjmp_on_fatal_error) + { + free_ret = 0; + ret = string + i + 2; + } + + for (t = 0; ret[t]; t++, j++) + temp[j] = ret[t]; + temp[j] = string[si]; + + if (string[si]) + { + j++; + i = si + 1; + } + else + i = si; + + if (free_ret) + free (ret); + continue; + } + + /* Add any character but a double quote to the quoted string we're + accumulating. */ + if (c != '"') + goto add_one_character; + + /* c == '"' */ + if (stripdq) + { + dquote ^= 1; + i++; + continue; + } + + break; + } + temp[j] = '\0'; + + /* Point to after the closing quote. */ + if (c) + i++; + *sindex = i; + + return (temp); +} + +/* This should really be another option to string_extract_double_quoted. */ +static int +skip_double_quoted (string, slen, sind) + char *string; + size_t slen; + int sind; +{ + int c, i; + char *ret; + int pass_next, backquote, si; + DECLARE_MBSTATE; + + pass_next = backquote = 0; + i = sind; + while (c = string[i]) + { + if (pass_next) + { + pass_next = 0; + ADVANCE_CHAR (string, slen, i); + continue; + } + else if (c == '\\') + { + pass_next++; + i++; + continue; + } + else if (backquote) + { + if (c == '`') + backquote = 0; + ADVANCE_CHAR (string, slen, i); + continue; + } + else if (c == '`') + { + backquote++; + i++; + continue; + } + else if (c == '$' && ((string[i + 1] == LPAREN) || (string[i + 1] == LBRACE))) + { + si = i + 2; + if (string[i + 1] == LPAREN) + ret = extract_command_subst (string, &si, SX_NOALLOC); + else + ret = extract_dollar_brace_string (string, &si, Q_DOUBLE_QUOTES, SX_NOALLOC); + + i = si + 1; + continue; + } + else if (c != '"') + { + ADVANCE_CHAR (string, slen, i); + continue; + } + else + break; + } + + if (c) + i++; + + return (i); +} + +/* Extract the contents of STRING as if it is enclosed in single quotes. + SINDEX, when passed in, is the offset of the character immediately + following the opening single quote; on exit, SINDEX is left pointing after + the closing single quote. */ +static inline char * +string_extract_single_quoted (string, sindex) + char *string; + int *sindex; +{ + register int i; + size_t slen; + char *t; + DECLARE_MBSTATE; + + /* Don't need slen for ADVANCE_CHAR unless multibyte chars possible. */ + slen = (MB_CUR_MAX > 1) ? strlen (string + *sindex) + *sindex : 0; + i = *sindex; + while (string[i] && string[i] != '\'') + ADVANCE_CHAR (string, slen, i); + + t = substring (string, *sindex, i); + + if (string[i]) + i++; + *sindex = i; + + return (t); +} + +static inline int +skip_single_quoted (string, slen, sind) + const char *string; + size_t slen; + int sind; +{ + register int c; + DECLARE_MBSTATE; + + c = sind; + while (string[c] && string[c] != '\'') + ADVANCE_CHAR (string, slen, c); + + if (string[c]) + c++; + return c; +} + +/* Just like string_extract, but doesn't hack backslashes or any of + that other stuff. Obeys CTLESC quoting. Used to do splitting on $IFS. */ +static char * +string_extract_verbatim (string, slen, sindex, charlist, flags) + char *string; + size_t slen; + int *sindex; + char *charlist; + int flags; +{ + register int i; +#if defined (HANDLE_MULTIBYTE) + size_t clen; + wchar_t *wcharlist; +#endif + int c; + char *temp; + DECLARE_MBSTATE; + + if (charlist[0] == '\'' && charlist[1] == '\0') + { + temp = string_extract_single_quoted (string, sindex); + --*sindex; /* leave *sindex at separator character */ + return temp; + } + + i = *sindex; +#if 0 + /* See how the MBLEN and ADVANCE_CHAR macros work to understand why we need + this only if MB_CUR_MAX > 1. */ + slen = (MB_CUR_MAX > 1) ? strlen (string + *sindex) + *sindex : 1; +#endif +#if defined (HANDLE_MULTIBYTE) + clen = strlen (charlist); + wcharlist = 0; +#endif + while (c = string[i]) + { +#if defined (HANDLE_MULTIBYTE) + size_t mblength; +#endif + if ((flags & SX_NOCTLESC) == 0 && c == CTLESC) + { + i += 2; + continue; + } + /* Even if flags contains SX_NOCTLESC, we let CTLESC quoting CTLNUL + through, to protect the CTLNULs from later calls to + remove_quoted_nulls. */ + else if ((flags & SX_NOESCCTLNUL) == 0 && c == CTLESC && string[i+1] == CTLNUL) + { + i += 2; + continue; + } + +#if defined (HANDLE_MULTIBYTE) + mblength = MBLEN (string + i, slen - i); + if (mblength > 1) + { + wchar_t wc; + mblength = mbtowc (&wc, string + i, slen - i); + if (MB_INVALIDCH (mblength)) + { + if (MEMBER (c, charlist)) + break; + } + else + { + if (wcharlist == 0) + { + size_t len; + len = mbstowcs (wcharlist, charlist, 0); + if (len == -1) + len = 0; + wcharlist = (wchar_t *)xmalloc (sizeof (wchar_t) * (len + 1)); + mbstowcs (wcharlist, charlist, len + 1); + } + + if (wcschr (wcharlist, wc)) + break; + } + } + else +#endif + if (MEMBER (c, charlist)) + break; + + ADVANCE_CHAR (string, slen, i); + } + +#if defined (HANDLE_MULTIBYTE) + FREE (wcharlist); +#endif + + temp = substring (string, *sindex, i); + *sindex = i; + + return (temp); +} + +/* Extract the $( construct in STRING, and return a new string. + Start extracting at (SINDEX) as if we had just seen "$(". + Make (SINDEX) get the position of the matching ")". ) + XFLAGS is additional flags to pass to other extraction functions. */ +char * +extract_command_subst (string, sindex, xflags) + char *string; + int *sindex; + int xflags; +{ + if (string[*sindex] == LPAREN) + return (extract_delimited_string (string, sindex, "$(", "(", ")", xflags|SX_COMMAND)); /*)*/ + else + { + xflags |= (no_longjmp_on_fatal_error ? SX_NOLONGJMP : 0); + return (xparse_dolparen (string, string+*sindex, sindex, xflags)); + } +} + +/* Extract the $[ construct in STRING, and return a new string. (]) + Start extracting at (SINDEX) as if we had just seen "$[". + Make (SINDEX) get the position of the matching "]". */ +char * +extract_arithmetic_subst (string, sindex) + char *string; + int *sindex; +{ + return (extract_delimited_string (string, sindex, "$[", "[", "]", 0)); /*]*/ +} + +#if defined (PROCESS_SUBSTITUTION) +/* Extract the <( or >( construct in STRING, and return a new string. + Start extracting at (SINDEX) as if we had just seen "<(". + Make (SINDEX) get the position of the matching ")". */ /*))*/ +char * +extract_process_subst (string, starter, sindex, xflags) + char *string; + char *starter; + int *sindex; + int xflags; +{ +#if 0 + return (extract_delimited_string (string, sindex, starter, "(", ")", SX_COMMAND)); +#else + xflags |= (no_longjmp_on_fatal_error ? SX_NOLONGJMP : 0); + return (xparse_dolparen (string, string+*sindex, sindex, xflags)); +#endif +} +#endif /* PROCESS_SUBSTITUTION */ + +#if defined (ARRAY_VARS) +/* This can be fooled by unquoted right parens in the passed string. If + each caller verifies that the last character in STRING is a right paren, + we don't even need to call extract_delimited_string. */ +char * +extract_array_assignment_list (string, sindex) + char *string; + int *sindex; +{ + int slen; + char *ret; + + slen = strlen (string); /* ( */ + if (string[slen - 1] == ')') + { + ret = substring (string, *sindex, slen - 1); + *sindex = slen - 1; + return ret; + } + return 0; +} +#endif + +/* Extract and create a new string from the contents of STRING, a + character string delimited with OPENER and CLOSER. SINDEX is + the address of an int describing the current offset in STRING; + it should point to just after the first OPENER found. On exit, + SINDEX gets the position of the last character of the matching CLOSER. + If OPENER is more than a single character, ALT_OPENER, if non-null, + contains a character string that can also match CLOSER and thus + needs to be skipped. */ +static char * +extract_delimited_string (string, sindex, opener, alt_opener, closer, flags) + char *string; + int *sindex; + char *opener, *alt_opener, *closer; + int flags; +{ + int i, c, si; + size_t slen; + char *t, *result; + int pass_character, nesting_level, in_comment; + int len_closer, len_opener, len_alt_opener; + DECLARE_MBSTATE; + + slen = strlen (string + *sindex) + *sindex; + len_opener = STRLEN (opener); + len_alt_opener = STRLEN (alt_opener); + len_closer = STRLEN (closer); + + pass_character = in_comment = 0; + + nesting_level = 1; + i = *sindex; + + while (nesting_level) + { + c = string[i]; + + if (c == 0) + break; + + if (in_comment) + { + if (c == '\n') + in_comment = 0; + ADVANCE_CHAR (string, slen, i); + continue; + } + + if (pass_character) /* previous char was backslash */ + { + pass_character = 0; + ADVANCE_CHAR (string, slen, i); + continue; + } + + /* Not exactly right yet; should handle shell metacharacters and + multibyte characters, too. See COMMENT_BEGIN define in parse.y */ + if ((flags & SX_COMMAND) && c == '#' && (i == 0 || string[i - 1] == '\n' || shellblank (string[i - 1]))) + { + in_comment = 1; + ADVANCE_CHAR (string, slen, i); + continue; + } + + if (c == CTLESC || c == '\\') + { + pass_character++; + i++; + continue; + } + + /* Process a nested command substitution, but only if we're parsing an + arithmetic substitution. */ + if ((flags & SX_COMMAND) && string[i] == '$' && string[i+1] == LPAREN) + { + si = i + 2; + t = extract_command_subst (string, &si, flags|SX_NOALLOC); + i = si + 1; + continue; + } + + /* Process a nested OPENER. */ + if (STREQN (string + i, opener, len_opener)) + { + si = i + len_opener; + t = extract_delimited_string (string, &si, opener, alt_opener, closer, flags|SX_NOALLOC); + i = si + 1; + continue; + } + + /* Process a nested ALT_OPENER */ + if (len_alt_opener && STREQN (string + i, alt_opener, len_alt_opener)) + { + si = i + len_alt_opener; + t = extract_delimited_string (string, &si, alt_opener, alt_opener, closer, flags|SX_NOALLOC); + i = si + 1; + continue; + } + + /* If the current substring terminates the delimited string, decrement + the nesting level. */ + if (STREQN (string + i, closer, len_closer)) + { + i += len_closer - 1; /* move to last byte of the closer */ + nesting_level--; + if (nesting_level == 0) + break; + } + + /* Pass old-style command substitution through verbatim. */ + if (c == '`') + { + si = i + 1; + t = string_extract (string, &si, "`", flags|SX_NOALLOC); + i = si + 1; + continue; + } + + /* Pass single-quoted and double-quoted strings through verbatim. */ + if (c == '\'' || c == '"') + { + si = i + 1; + i = (c == '\'') ? skip_single_quoted (string, slen, si) + : skip_double_quoted (string, slen, si); + continue; + } + + /* move past this character, which was not special. */ + ADVANCE_CHAR (string, slen, i); + } + + if (c == 0 && nesting_level) + { + if (no_longjmp_on_fatal_error == 0) + { + last_command_exit_value = EXECUTION_FAILURE; + report_error (_("bad substitution: no closing `%s' in %s"), closer, string); + exp_jump_to_top_level (DISCARD); + } + else + { + *sindex = i; + return (char *)NULL; + } + } + + si = i - *sindex - len_closer + 1; + if (flags & SX_NOALLOC) + result = (char *)NULL; + else + { + result = (char *)xmalloc (1 + si); + strncpy (result, string + *sindex, si); + result[si] = '\0'; + } + *sindex = i; + + return (result); +} + +/* Extract a parameter expansion expression within ${ and } from STRING. + Obey the Posix.2 rules for finding the ending `}': count braces while + skipping over enclosed quoted strings and command substitutions. + SINDEX is the address of an int describing the current offset in STRING; + it should point to just after the first `{' found. On exit, SINDEX + gets the position of the matching `}'. QUOTED is non-zero if this + occurs inside double quotes. */ +/* XXX -- this is very similar to extract_delimited_string -- XXX */ +static char * +extract_dollar_brace_string (string, sindex, quoted, flags) + char *string; + int *sindex, quoted, flags; +{ + register int i, c; + size_t slen; + int pass_character, nesting_level, si, dolbrace_state; + char *result, *t; + DECLARE_MBSTATE; + + pass_character = 0; + nesting_level = 1; + slen = strlen (string + *sindex) + *sindex; + + /* The handling of dolbrace_state needs to agree with the code in parse.y: + parse_matched_pair(). The different initial value is to handle the + case where this function is called to parse the word in + ${param op word} (SX_WORD). */ + dolbrace_state = (flags & SX_WORD) ? DOLBRACE_WORD : DOLBRACE_PARAM; + if ((quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) && (flags & SX_POSIXEXP)) + dolbrace_state = DOLBRACE_QUOTE; + + i = *sindex; + while (c = string[i]) + { + if (pass_character) + { + pass_character = 0; + ADVANCE_CHAR (string, slen, i); + continue; + } + + /* CTLESCs and backslashes quote the next character. */ + if (c == CTLESC || c == '\\') + { + pass_character++; + i++; + continue; + } + + if (string[i] == '$' && string[i+1] == LBRACE) + { + nesting_level++; + i += 2; + continue; + } + + if (c == RBRACE) + { + nesting_level--; + if (nesting_level == 0) + break; + i++; + continue; + } + + /* Pass the contents of old-style command substitutions through + verbatim. */ + if (c == '`') + { + si = i + 1; + t = string_extract (string, &si, "`", flags|SX_NOALLOC); + i = si + 1; + continue; + } + + /* Pass the contents of new-style command substitutions and + arithmetic substitutions through verbatim. */ + if (string[i] == '$' && string[i+1] == LPAREN) + { + si = i + 2; + t = extract_command_subst (string, &si, flags|SX_NOALLOC); + i = si + 1; + continue; + } + + /* Pass the contents of double-quoted strings through verbatim. */ + if (c == '"') + { + si = i + 1; + i = skip_double_quoted (string, slen, si); + /* skip_XXX_quoted leaves index one past close quote */ + continue; + } + + if (c == '\'') + { +/*itrace("extract_dollar_brace_string: c == single quote flags = %d quoted = %d dolbrace_state = %d", flags, quoted, dolbrace_state);*/ + if (posixly_correct && shell_compatibility_level > 42 && dolbrace_state != DOLBRACE_QUOTE && (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES))) + ADVANCE_CHAR (string, slen, i); + else + { + si = i + 1; + i = skip_single_quoted (string, slen, si); + } + + continue; + } + + /* move past this character, which was not special. */ + ADVANCE_CHAR (string, slen, i); + + /* This logic must agree with parse.y:parse_matched_pair, since they + share the same defines. */ + if (dolbrace_state == DOLBRACE_PARAM && c == '%' && (i - *sindex) > 1) + dolbrace_state = DOLBRACE_QUOTE; + else if (dolbrace_state == DOLBRACE_PARAM && c == '#' && (i - *sindex) > 1) + dolbrace_state = DOLBRACE_QUOTE; + else if (dolbrace_state == DOLBRACE_PARAM && c == '/' && (i - *sindex) > 1) + dolbrace_state = DOLBRACE_QUOTE2; /* XXX */ + else if (dolbrace_state == DOLBRACE_PARAM && c == '^' && (i - *sindex) > 1) + dolbrace_state = DOLBRACE_QUOTE; + else if (dolbrace_state == DOLBRACE_PARAM && c == ',' && (i - *sindex) > 1) + dolbrace_state = DOLBRACE_QUOTE; + else if (dolbrace_state == DOLBRACE_PARAM && strchr ("#%^,~:-=?+/", c) != 0) + dolbrace_state = DOLBRACE_OP; + else if (dolbrace_state == DOLBRACE_OP && strchr ("#%^,~:-=?+/", c) == 0) + dolbrace_state = DOLBRACE_WORD; + } + + if (c == 0 && nesting_level) + { + if (no_longjmp_on_fatal_error == 0) + { /* { */ + last_command_exit_value = EXECUTION_FAILURE; + report_error (_("bad substitution: no closing `%s' in %s"), "}", string); + exp_jump_to_top_level (DISCARD); + } + else + { + *sindex = i; + return ((char *)NULL); + } + } + + result = (flags & SX_NOALLOC) ? (char *)NULL : substring (string, *sindex, i); + *sindex = i; + + return (result); +} + +/* Remove backslashes which are quoting backquotes from STRING. Modifies + STRING, and returns a pointer to it. */ +char * +de_backslash (string) + char *string; +{ + register size_t slen; + register int i, j, prev_i; + DECLARE_MBSTATE; + + slen = strlen (string); + i = j = 0; + + /* Loop copying string[i] to string[j], i >= j. */ + while (i < slen) + { + if (string[i] == '\\' && (string[i + 1] == '`' || string[i + 1] == '\\' || + string[i + 1] == '$')) + i++; + prev_i = i; + ADVANCE_CHAR (string, slen, i); + if (j < prev_i) + do string[j++] = string[prev_i++]; while (prev_i < i); + else + j = i; + } + string[j] = '\0'; + + return (string); +} + +#if 0 +/*UNUSED*/ +/* Replace instances of \! in a string with !. */ +void +unquote_bang (string) + char *string; +{ + register int i, j; + register char *temp; + + temp = (char *)xmalloc (1 + strlen (string)); + + for (i = 0, j = 0; (temp[j] = string[i]); i++, j++) + { + if (string[i] == '\\' && string[i + 1] == '!') + { + temp[j] = '!'; + i++; + } + } + strcpy (string, temp); + free (temp); +} +#endif + +#define CQ_RETURN(x) do { no_longjmp_on_fatal_error = 0; return (x); } while (0) + +/* This function assumes s[i] == open; returns with s[ret] == close; used to + parse array subscripts. FLAGS & 1 means to not attempt to skip over + matched pairs of quotes or backquotes, or skip word expansions; it is + intended to be used after expansion has been performed and during final + assignment parsing (see arrayfunc.c:assign_compound_array_list()). */ +static int +skip_matched_pair (string, start, open, close, flags) + const char *string; + int start, open, close, flags; +{ + int i, pass_next, backq, si, c, count; + size_t slen; + char *temp, *ss; + DECLARE_MBSTATE; + + slen = strlen (string + start) + start; + no_longjmp_on_fatal_error = 1; + + i = start + 1; /* skip over leading bracket */ + count = 1; + pass_next = backq = 0; + ss = (char *)string; + while (c = string[i]) + { + if (pass_next) + { + pass_next = 0; + if (c == 0) + CQ_RETURN(i); + ADVANCE_CHAR (string, slen, i); + continue; + } + else if (c == '\\') + { + pass_next = 1; + i++; + continue; + } + else if (backq) + { + if (c == '`') + backq = 0; + ADVANCE_CHAR (string, slen, i); + continue; + } + else if ((flags & 1) == 0 && c == '`') + { + backq = 1; + i++; + continue; + } + else if ((flags & 1) == 0 && c == open) + { + count++; + i++; + continue; + } + else if (c == close) + { + count--; + if (count == 0) + break; + i++; + continue; + } + else if ((flags & 1) == 0 && (c == '\'' || c == '"')) + { + i = (c == '\'') ? skip_single_quoted (ss, slen, ++i) + : skip_double_quoted (ss, slen, ++i); + /* no increment, the skip functions increment past the closing quote. */ + } + else if ((flags&1) == 0 && c == '$' && (string[i+1] == LPAREN || string[i+1] == LBRACE)) + { + si = i + 2; + if (string[si] == '\0') + CQ_RETURN(si); + + if (string[i+1] == LPAREN) + temp = extract_delimited_string (ss, &si, "$(", "(", ")", SX_NOALLOC|SX_COMMAND); /* ) */ + else + temp = extract_dollar_brace_string (ss, &si, 0, SX_NOALLOC); + i = si; + if (string[i] == '\0') /* don't increment i past EOS in loop */ + break; + i++; + continue; + } + else + ADVANCE_CHAR (string, slen, i); + } + + CQ_RETURN(i); +} + +#if defined (ARRAY_VARS) +int +skipsubscript (string, start, flags) + const char *string; + int start, flags; +{ + return (skip_matched_pair (string, start, '[', ']', flags)); +} +#endif + +/* Skip characters in STRING until we find a character in DELIMS, and return + the index of that character. START is the index into string at which we + begin. This is similar in spirit to strpbrk, but it returns an index into + STRING and takes a starting index. This little piece of code knows quite + a lot of shell syntax. It's very similar to skip_double_quoted and other + functions of that ilk. */ +int +skip_to_delim (string, start, delims, flags) + char *string; + int start; + char *delims; + int flags; +{ + int i, pass_next, backq, si, c, invert, skipquote, skipcmd, noprocsub; + size_t slen; + char *temp, open[3]; + DECLARE_MBSTATE; + + slen = strlen (string + start) + start; + if (flags & SD_NOJMP) + no_longjmp_on_fatal_error = 1; + invert = (flags & SD_INVERT); + skipcmd = (flags & SD_NOSKIPCMD) == 0; + noprocsub = (flags & SD_NOPROCSUB); + + i = start; + pass_next = backq = 0; + while (c = string[i]) + { + /* If this is non-zero, we should not let quote characters be delimiters + and the current character is a single or double quote. We should not + test whether or not it's a delimiter until after we skip single- or + double-quoted strings. */ + skipquote = ((flags & SD_NOQUOTEDELIM) && (c == '\'' || c =='"')); + if (pass_next) + { + pass_next = 0; + if (c == 0) + CQ_RETURN(i); + ADVANCE_CHAR (string, slen, i); + continue; + } + else if (c == '\\') + { + pass_next = 1; + i++; + continue; + } + else if (backq) + { + if (c == '`') + backq = 0; + ADVANCE_CHAR (string, slen, i); + continue; + } + else if (c == '`') + { + backq = 1; + i++; + continue; + } + else if (skipquote == 0 && invert == 0 && member (c, delims)) + break; + else if (c == '\'' || c == '"') + { + i = (c == '\'') ? skip_single_quoted (string, slen, ++i) + : skip_double_quoted (string, slen, ++i); + /* no increment, the skip functions increment past the closing quote. */ + } + else if (c == '$' && ((skipcmd && string[i+1] == LPAREN) || string[i+1] == LBRACE)) + { + si = i + 2; + if (string[si] == '\0') + CQ_RETURN(si); + + if (string[i+1] == LPAREN) + temp = extract_delimited_string (string, &si, "$(", "(", ")", SX_NOALLOC|SX_COMMAND); /* ) */ + else + temp = extract_dollar_brace_string (string, &si, 0, SX_NOALLOC); + i = si; + if (string[i] == '\0') /* don't increment i past EOS in loop */ + break; + i++; + continue; + } +#if defined (PROCESS_SUBSTITUTION) + else if (skipcmd && noprocsub == 0 && (c == '<' || c == '>') && string[i+1] == LPAREN) + { + si = i + 2; + if (string[si] == '\0') + CQ_RETURN(si); + temp = extract_process_subst (string, (c == '<') ? "<(" : ">(", &si, 0); + free (temp); /* XXX - not using SX_ALLOC here yet */ + i = si; + if (string[i] == '\0') + break; + i++; + continue; + } +#endif /* PROCESS_SUBSTITUTION */ +#if defined (EXTENDED_GLOB) + else if ((flags & SD_EXTGLOB) && extended_glob && string[i+1] == LPAREN && member (c, "?*+!@")) + { + si = i + 2; + if (string[si] == '\0') + CQ_RETURN(si); + + open[0] = c; + open[1] = LPAREN; + open[2] = '\0'; + temp = extract_delimited_string (string, &si, open, "(", ")", SX_NOALLOC); /* ) */ + + i = si; + if (string[i] == '\0') /* don't increment i past EOS in loop */ + break; + i++; + continue; + } +#endif + else if ((flags & SD_GLOB) && c == LBRACK) + { + si = i + 1; + if (string[si] == '\0') + CQ_RETURN(si); + + temp = extract_delimited_string (string, &si, "[", "[", "]", SX_NOALLOC); /* ] */ + + i = si; + if (string[i] == '\0') /* don't increment i past EOS in loop */ + break; + i++; + continue; + } + else if ((skipquote || invert) && (member (c, delims) == 0)) + break; + else + ADVANCE_CHAR (string, slen, i); + } + + CQ_RETURN(i); +} + +#if defined (READLINE) +/* Return 1 if the portion of STRING ending at EINDEX is quoted (there is + an unclosed quoted string), or if the character at EINDEX is quoted + by a backslash. NO_LONGJMP_ON_FATAL_ERROR is used to flag that the various + single and double-quoted string parsing functions should not return an + error if there are unclosed quotes or braces. The characters that this + recognizes need to be the same as the contents of + rl_completer_quote_characters. */ + +int +char_is_quoted (string, eindex) + char *string; + int eindex; +{ + int i, pass_next, c; + size_t slen; + DECLARE_MBSTATE; + + slen = strlen (string); + no_longjmp_on_fatal_error = 1; + i = pass_next = 0; + while (i <= eindex) + { + c = string[i]; + + if (pass_next) + { + pass_next = 0; + if (i >= eindex) /* XXX was if (i >= eindex - 1) */ + CQ_RETURN(1); + ADVANCE_CHAR (string, slen, i); + continue; + } + else if (c == '\\') + { + pass_next = 1; + i++; + continue; + } + else if (c == '\'' || c == '"') + { + i = (c == '\'') ? skip_single_quoted (string, slen, ++i) + : skip_double_quoted (string, slen, ++i); + if (i > eindex) + CQ_RETURN(1); + /* no increment, the skip_xxx functions go one past end */ + } + else + ADVANCE_CHAR (string, slen, i); + } + + CQ_RETURN(0); +} + +int +unclosed_pair (string, eindex, openstr) + char *string; + int eindex; + char *openstr; +{ + int i, pass_next, openc, olen; + size_t slen; + DECLARE_MBSTATE; + + slen = strlen (string); + olen = strlen (openstr); + i = pass_next = openc = 0; + while (i <= eindex) + { + if (pass_next) + { + pass_next = 0; + if (i >= eindex) /* XXX was if (i >= eindex - 1) */ + return 0; + ADVANCE_CHAR (string, slen, i); + continue; + } + else if (string[i] == '\\') + { + pass_next = 1; + i++; + continue; + } + else if (STREQN (string + i, openstr, olen)) + { + openc = 1 - openc; + i += olen; + } + else if (string[i] == '\'' || string[i] == '"') + { + i = (string[i] == '\'') ? skip_single_quoted (string, slen, i) + : skip_double_quoted (string, slen, i); + if (i > eindex) + return 0; + } + else + ADVANCE_CHAR (string, slen, i); + } + return (openc); +} + +/* Split STRING (length SLEN) at DELIMS, and return a WORD_LIST with the + individual words. If DELIMS is NULL, the current value of $IFS is used + to split the string, and the function follows the shell field splitting + rules. SENTINEL is an index to look for. NWP, if non-NULL, + gets the number of words in the returned list. CWP, if non-NULL, gets + the index of the word containing SENTINEL. Non-whitespace chars in + DELIMS delimit separate fields. */ +WORD_LIST * +split_at_delims (string, slen, delims, sentinel, flags, nwp, cwp) + char *string; + int slen; + char *delims; + int sentinel, flags; + int *nwp, *cwp; +{ + int ts, te, i, nw, cw, ifs_split, dflags; + char *token, *d, *d2; + WORD_LIST *ret, *tl; + + if (string == 0 || *string == '\0') + { + if (nwp) + *nwp = 0; + if (cwp) + *cwp = 0; + return ((WORD_LIST *)NULL); + } + + d = (delims == 0) ? ifs_value : delims; + ifs_split = delims == 0; + + /* Make d2 the non-whitespace characters in delims */ + d2 = 0; + if (delims) + { + size_t slength; +#if defined (HANDLE_MULTIBYTE) + size_t mblength = 1; +#endif + DECLARE_MBSTATE; + + slength = strlen (delims); + d2 = (char *)xmalloc (slength + 1); + i = ts = 0; + while (delims[i]) + { +#if defined (HANDLE_MULTIBYTE) + mbstate_t state_bak; + state_bak = state; + mblength = MBRLEN (delims + i, slength, &state); + if (MB_INVALIDCH (mblength)) + state = state_bak; + else if (mblength > 1) + { + memcpy (d2 + ts, delims + i, mblength); + ts += mblength; + i += mblength; + slength -= mblength; + continue; + } +#endif + if (whitespace (delims[i]) == 0) + d2[ts++] = delims[i]; + + i++; + slength--; + } + d2[ts] = '\0'; + } + + ret = (WORD_LIST *)NULL; + + /* Remove sequences of whitespace characters at the start of the string, as + long as those characters are delimiters. */ + for (i = 0; member (string[i], d) && spctabnl (string[i]); i++) + ; + if (string[i] == '\0') + return (ret); + + ts = i; + nw = 0; + cw = -1; + dflags = flags|SD_NOJMP; + while (1) + { + te = skip_to_delim (string, ts, d, dflags); + + /* If we have a non-whitespace delimiter character, use it to make a + separate field. This is just about what $IFS splitting does and + is closer to the behavior of the shell parser. */ + if (ts == te && d2 && member (string[ts], d2)) + { + te = ts + 1; + /* If we're using IFS splitting, the non-whitespace delimiter char + and any additional IFS whitespace delimits a field. */ + if (ifs_split) + while (member (string[te], d) && spctabnl (string[te])) + te++; + else + while (member (string[te], d2)) + te++; + } + + token = substring (string, ts, te); + + ret = add_string_to_list (token, ret); + free (token); + nw++; + + if (sentinel >= ts && sentinel <= te) + cw = nw; + + /* If the cursor is at whitespace just before word start, set the + sentinel word to the current word. */ + if (cwp && cw == -1 && sentinel == ts-1) + cw = nw; + + /* If the cursor is at whitespace between two words, make a new, empty + word, add it before (well, after, since the list is in reverse order) + the word we just added, and set the current word to that one. */ + if (cwp && cw == -1 && sentinel < ts) + { + tl = make_word_list (make_word (""), ret->next); + ret->next = tl; + cw = nw; + nw++; + } + + if (string[te] == 0) + break; + + i = te; + while (member (string[i], d) && (ifs_split || spctabnl(string[i]))) + i++; + + if (string[i]) + ts = i; + else + break; + } + + /* Special case for SENTINEL at the end of STRING. If we haven't found + the word containing SENTINEL yet, and the index we're looking for is at + the end of STRING (or past the end of the previously-found token, + possible if the end of the line is composed solely of IFS whitespace) + add an additional null argument and set the current word pointer to that. */ + if (cwp && cw == -1 && (sentinel >= slen || sentinel >= te)) + { + if (whitespace (string[sentinel - 1])) + { + token = ""; + ret = add_string_to_list (token, ret); + nw++; + } + cw = nw; + } + + if (nwp) + *nwp = nw; + if (cwp) + *cwp = cw; + + FREE (d2); + + return (REVERSE_LIST (ret, WORD_LIST *)); +} +#endif /* READLINE */ + +#if 0 +/* UNUSED */ +/* Extract the name of the variable to bind to from the assignment string. */ +char * +assignment_name (string) + char *string; +{ + int offset; + char *temp; + + offset = assignment (string, 0); + if (offset == 0) + return (char *)NULL; + temp = substring (string, 0, offset); + return (temp); +} +#endif + +/* **************************************************************** */ +/* */ +/* Functions to convert strings to WORD_LISTs and vice versa */ +/* */ +/* **************************************************************** */ + +/* Return a single string of all the words in LIST. SEP is the separator + to put between individual elements of LIST in the output string. */ +char * +string_list_internal (list, sep) + WORD_LIST *list; + char *sep; +{ + register WORD_LIST *t; + char *result, *r; + int word_len, sep_len, result_size; + + if (list == 0) + return ((char *)NULL); + + /* Short-circuit quickly if we don't need to separate anything. */ + if (list->next == 0) + return (savestring (list->word->word)); + + /* This is nearly always called with either sep[0] == 0 or sep[1] == 0. */ + sep_len = STRLEN (sep); + result_size = 0; + + for (t = list; t; t = t->next) + { + if (t != list) + result_size += sep_len; + result_size += strlen (t->word->word); + } + + r = result = (char *)xmalloc (result_size + 1); + + for (t = list; t; t = t->next) + { + if (t != list && sep_len) + { + if (sep_len > 1) + { + FASTCOPY (sep, r, sep_len); + r += sep_len; + } + else + *r++ = sep[0]; + } + + word_len = strlen (t->word->word); + FASTCOPY (t->word->word, r, word_len); + r += word_len; + } + + *r = '\0'; + return (result); +} + +/* Return a single string of all the words present in LIST, separating + each word with a space. */ +char * +string_list (list) + WORD_LIST *list; +{ + return (string_list_internal (list, " ")); +} + +/* An external interface that can be used by the rest of the shell to + obtain a string containing the first character in $IFS. Handles all + the multibyte complications. If LENP is non-null, it is set to the + length of the returned string. */ +char * +ifs_firstchar (lenp) + int *lenp; +{ + char *ret; + int len; + + ret = xmalloc (MB_LEN_MAX + 1); +#if defined (HANDLE_MULTIBYTE) + if (ifs_firstc_len == 1) + { + ret[0] = ifs_firstc[0]; + ret[1] = '\0'; + len = ret[0] ? 1 : 0; + } + else + { + memcpy (ret, ifs_firstc, ifs_firstc_len); + ret[len = ifs_firstc_len] = '\0'; + } +#else + ret[0] = ifs_firstc; + ret[1] = '\0'; + len = ret[0] ? 0 : 1; +#endif + + if (lenp) + *lenp = len; + + return ret; +} + +/* Return a single string of all the words present in LIST, obeying the + quoting rules for "$*", to wit: (P1003.2, draft 11, 3.5.2) "If the + expansion [of $*] appears within a double quoted string, it expands + to a single field with the value of each parameter separated by the + first character of the IFS variable, or by a if IFS is unset." */ +char * +string_list_dollar_star (list) + WORD_LIST *list; +{ + char *ret; +#if defined (HANDLE_MULTIBYTE) +# if defined (__GNUC__) + char sep[MB_CUR_MAX + 1]; +# else + char *sep = 0; +# endif +#else + char sep[2]; +#endif + +#if defined (HANDLE_MULTIBYTE) +# if !defined (__GNUC__) + sep = (char *)xmalloc (MB_CUR_MAX + 1); +# endif /* !__GNUC__ */ + if (ifs_firstc_len == 1) + { + sep[0] = ifs_firstc[0]; + sep[1] = '\0'; + } + else + { + memcpy (sep, ifs_firstc, ifs_firstc_len); + sep[ifs_firstc_len] = '\0'; + } +#else + sep[0] = ifs_firstc; + sep[1] = '\0'; +#endif + + ret = string_list_internal (list, sep); +#if defined (HANDLE_MULTIBYTE) && !defined (__GNUC__) + free (sep); +#endif + return ret; +} + +/* Turn $@ into a string. If (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) + is non-zero, the $@ appears within double quotes, and we should quote + the list before converting it into a string. If IFS is unset, and the + word is not quoted, we just need to quote CTLESC and CTLNUL characters + in the words in the list, because the default value of $IFS is + , IFS characters in the words in the list should + also be split. If IFS is null, and the word is not quoted, we need + to quote the words in the list to preserve the positional parameters + exactly. */ +char * +string_list_dollar_at (list, quoted) + WORD_LIST *list; + int quoted; +{ + char *ifs, *ret; +#if defined (HANDLE_MULTIBYTE) +# if defined (__GNUC__) + char sep[MB_CUR_MAX + 1]; +# else + char *sep = 0; +# endif /* !__GNUC__ */ +#else + char sep[2]; +#endif + WORD_LIST *tlist; + + /* XXX this could just be ifs = ifs_value; */ + ifs = ifs_var ? value_cell (ifs_var) : (char *)0; + +#if defined (HANDLE_MULTIBYTE) +# if !defined (__GNUC__) + sep = (char *)xmalloc (MB_CUR_MAX + 1); +# endif /* !__GNUC__ */ + if (ifs && *ifs) + { + if (ifs_firstc_len == 1) + { + sep[0] = ifs_firstc[0]; + sep[1] = '\0'; + } + else + { + memcpy (sep, ifs_firstc, ifs_firstc_len); + sep[ifs_firstc_len] = '\0'; + } + } + else + { + sep[0] = ' '; + sep[1] = '\0'; + } +#else + sep[0] = (ifs == 0 || *ifs == 0) ? ' ' : *ifs; + sep[1] = '\0'; +#endif + + /* XXX -- why call quote_list if ifs == 0? we can get away without doing + it now that quote_escapes quotes spaces */ + tlist = (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES|Q_PATQUOTE)) + ? quote_list (list) + : list_quote_escapes (list); + + ret = string_list_internal (tlist, sep); +#if defined (HANDLE_MULTIBYTE) && !defined (__GNUC__) + free (sep); +#endif + return ret; +} + +/* Turn the positional parameters into a string, understanding quoting and + the various subtleties of using the first character of $IFS as the + separator. Calls string_list_dollar_at, string_list_dollar_star, and + string_list as appropriate. */ +char * +string_list_pos_params (pchar, list, quoted) + int pchar; + WORD_LIST *list; + int quoted; +{ + char *ret; + WORD_LIST *tlist; + + if (pchar == '*' && (quoted & Q_DOUBLE_QUOTES)) + { + tlist = quote_list (list); + word_list_remove_quoted_nulls (tlist); + ret = string_list_dollar_star (tlist); + } + else if (pchar == '*' && (quoted & Q_HERE_DOCUMENT)) + { + tlist = quote_list (list); + word_list_remove_quoted_nulls (tlist); + ret = string_list (tlist); + } + else if (pchar == '*') + { + /* Even when unquoted, string_list_dollar_star does the right thing + making sure that the first character of $IFS is used as the + separator. */ + ret = string_list_dollar_star (list); + } + else if (pchar == '@' && (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES))) + /* We use string_list_dollar_at, but only if the string is quoted, since + that quotes the escapes if it's not, which we don't want. We could + use string_list (the old code did), but that doesn't do the right + thing if the first character of $IFS is not a space. We use + string_list_dollar_star if the string is unquoted so we make sure that + the elements of $@ are separated by the first character of $IFS for + later splitting. */ + ret = string_list_dollar_at (list, quoted); + else if (pchar == '@') + ret = string_list_dollar_star (list); + else + ret = string_list ((quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) ? quote_list (list) : list); + + return ret; +} + +/* Return the list of words present in STRING. Separate the string into + words at any of the characters found in SEPARATORS. If QUOTED is + non-zero then word in the list will have its quoted flag set, otherwise + the quoted flag is left as make_word () deemed fit. + + This obeys the P1003.2 word splitting semantics. If `separators' is + exactly , then the splitting algorithm is that of + the Bourne shell, which treats any sequence of characters from `separators' + as a delimiter. If IFS is unset, which results in `separators' being set + to "", no splitting occurs. If separators has some other value, the + following rules are applied (`IFS white space' means zero or more + occurrences of , , or , as long as those characters + are in `separators'): + + 1) IFS white space is ignored at the start and the end of the + string. + 2) Each occurrence of a character in `separators' that is not + IFS white space, along with any adjacent occurrences of + IFS white space delimits a field. + 3) Any nonzero-length sequence of IFS white space delimits a field. + */ + +/* BEWARE! list_string strips null arguments. Don't call it twice and + expect to have "" preserved! */ + +/* This performs word splitting and quoted null character removal on + STRING. */ +#define issep(c) \ + (((separators)[0]) ? ((separators)[1] ? isifs(c) \ + : (c) == (separators)[0]) \ + : 0) + +WORD_LIST * +list_string (string, separators, quoted) + register char *string, *separators; + int quoted; +{ + WORD_LIST *result; + WORD_DESC *t; + char *current_word, *s; + int sindex, sh_style_split, whitesep, xflags; + size_t slen; + + if (!string || !*string) + return ((WORD_LIST *)NULL); + + sh_style_split = separators && separators[0] == ' ' && + separators[1] == '\t' && + separators[2] == '\n' && + separators[3] == '\0'; + for (xflags = 0, s = ifs_value; s && *s; s++) + { + if (*s == CTLESC) xflags |= SX_NOCTLESC; + else if (*s == CTLNUL) xflags |= SX_NOESCCTLNUL; + } + + slen = 0; + /* Remove sequences of whitespace at the beginning of STRING, as + long as those characters appear in IFS. Do not do this if + STRING is quoted or if there are no separator characters. */ + if (!quoted || !separators || !*separators) + { + for (s = string; *s && spctabnl (*s) && issep (*s); s++); + + if (!*s) + return ((WORD_LIST *)NULL); + + string = s; + } + + /* OK, now STRING points to a word that does not begin with white space. + The splitting algorithm is: + extract a word, stopping at a separator + skip sequences of spc, tab, or nl as long as they are separators + This obeys the field splitting rules in Posix.2. */ + slen = (MB_CUR_MAX > 1) ? strlen (string) : 1; + for (result = (WORD_LIST *)NULL, sindex = 0; string[sindex]; ) + { + /* Don't need string length in ADVANCE_CHAR or string_extract_verbatim + unless multibyte chars are possible. */ + current_word = string_extract_verbatim (string, slen, &sindex, separators, xflags); + if (current_word == 0) + break; + + /* If we have a quoted empty string, add a quoted null argument. We + want to preserve the quoted null character iff this is a quoted + empty string; otherwise the quoted null characters are removed + below. */ + if (QUOTED_NULL (current_word)) + { + t = alloc_word_desc (); + t->word = make_quoted_char ('\0'); + t->flags |= W_QUOTED|W_HASQUOTEDNULL; + result = make_word_list (t, result); + } + else if (current_word[0] != '\0') + { + /* If we have something, then add it regardless. However, + perform quoted null character removal on the current word. */ + remove_quoted_nulls (current_word); + result = add_string_to_list (current_word, result); + result->word->flags &= ~W_HASQUOTEDNULL; /* just to be sure */ + if (quoted & (Q_DOUBLE_QUOTES|Q_HERE_DOCUMENT)) + result->word->flags |= W_QUOTED; + } + + /* If we're not doing sequences of separators in the traditional + Bourne shell style, then add a quoted null argument. */ + else if (!sh_style_split && !spctabnl (string[sindex])) + { + t = alloc_word_desc (); + t->word = make_quoted_char ('\0'); + t->flags |= W_QUOTED|W_HASQUOTEDNULL; + result = make_word_list (t, result); + } + + free (current_word); + + /* Note whether or not the separator is IFS whitespace, used later. */ + whitesep = string[sindex] && spctabnl (string[sindex]); + + /* Move past the current separator character. */ + if (string[sindex]) + { + DECLARE_MBSTATE; + ADVANCE_CHAR (string, slen, sindex); + } + + /* Now skip sequences of space, tab, or newline characters if they are + in the list of separators. */ + while (string[sindex] && spctabnl (string[sindex]) && issep (string[sindex])) + sindex++; + + /* If the first separator was IFS whitespace and the current character + is a non-whitespace IFS character, it should be part of the current + field delimiter, not a separate delimiter that would result in an + empty field. Look at POSIX.2, 3.6.5, (3)(b). */ + if (string[sindex] && whitesep && issep (string[sindex]) && !spctabnl (string[sindex])) + { + sindex++; + /* An IFS character that is not IFS white space, along with any + adjacent IFS white space, shall delimit a field. (SUSv3) */ + while (string[sindex] && spctabnl (string[sindex]) && isifs (string[sindex])) + sindex++; + } + } + return (REVERSE_LIST (result, WORD_LIST *)); +} + +/* Parse a single word from STRING, using SEPARATORS to separate fields. + ENDPTR is set to the first character after the word. This is used by + the `read' builtin. This is never called with SEPARATORS != $IFS; + it should be simplified. + + XXX - this function is very similar to list_string; they should be + combined - XXX */ +char * +get_word_from_string (stringp, separators, endptr) + char **stringp, *separators, **endptr; +{ + register char *s; + char *current_word; + int sindex, sh_style_split, whitesep, xflags; + size_t slen; + + if (!stringp || !*stringp || !**stringp) + return ((char *)NULL); + + sh_style_split = separators && separators[0] == ' ' && + separators[1] == '\t' && + separators[2] == '\n' && + separators[3] == '\0'; + for (xflags = 0, s = ifs_value; s && *s; s++) + { + if (*s == CTLESC) xflags |= SX_NOCTLESC; + if (*s == CTLNUL) xflags |= SX_NOESCCTLNUL; + } + + s = *stringp; + slen = 0; + + /* Remove sequences of whitespace at the beginning of STRING, as + long as those characters appear in IFS. */ + if (sh_style_split || !separators || !*separators) + { + for (; *s && spctabnl (*s) && isifs (*s); s++); + + /* If the string is nothing but whitespace, update it and return. */ + if (!*s) + { + *stringp = s; + if (endptr) + *endptr = s; + return ((char *)NULL); + } + } + + /* OK, S points to a word that does not begin with white space. + Now extract a word, stopping at a separator, save a pointer to + the first character after the word, then skip sequences of spc, + tab, or nl as long as they are separators. + + This obeys the field splitting rules in Posix.2. */ + sindex = 0; + /* Don't need string length in ADVANCE_CHAR or string_extract_verbatim + unless multibyte chars are possible. */ + slen = (MB_CUR_MAX > 1) ? strlen (s) : 1; + current_word = string_extract_verbatim (s, slen, &sindex, separators, xflags); + + /* Set ENDPTR to the first character after the end of the word. */ + if (endptr) + *endptr = s + sindex; + + /* Note whether or not the separator is IFS whitespace, used later. */ + whitesep = s[sindex] && spctabnl (s[sindex]); + + /* Move past the current separator character. */ + if (s[sindex]) + { + DECLARE_MBSTATE; + ADVANCE_CHAR (s, slen, sindex); + } + + /* Now skip sequences of space, tab, or newline characters if they are + in the list of separators. */ + while (s[sindex] && spctabnl (s[sindex]) && isifs (s[sindex])) + sindex++; + + /* If the first separator was IFS whitespace and the current character is + a non-whitespace IFS character, it should be part of the current field + delimiter, not a separate delimiter that would result in an empty field. + Look at POSIX.2, 3.6.5, (3)(b). */ + if (s[sindex] && whitesep && isifs (s[sindex]) && !spctabnl (s[sindex])) + { + sindex++; + /* An IFS character that is not IFS white space, along with any adjacent + IFS white space, shall delimit a field. */ + while (s[sindex] && spctabnl (s[sindex]) && isifs (s[sindex])) + sindex++; + } + + /* Update STRING to point to the next field. */ + *stringp = s + sindex; + return (current_word); +} + +/* Remove IFS white space at the end of STRING. Start at the end + of the string and walk backwards until the beginning of the string + or we find a character that's not IFS white space and not CTLESC. + Only let CTLESC escape a white space character if SAW_ESCAPE is + non-zero. */ +char * +strip_trailing_ifs_whitespace (string, separators, saw_escape) + char *string, *separators; + int saw_escape; +{ + char *s; + + s = string + STRLEN (string) - 1; + while (s > string && ((spctabnl (*s) && isifs (*s)) || + (saw_escape && *s == CTLESC && spctabnl (s[1])))) + s--; + *++s = '\0'; + return string; +} + +#if 0 +/* UNUSED */ +/* Split STRING into words at whitespace. Obeys shell-style quoting with + backslashes, single and double quotes. */ +WORD_LIST * +list_string_with_quotes (string) + char *string; +{ + WORD_LIST *list; + char *token, *s; + size_t s_len; + int c, i, tokstart, len; + + for (s = string; s && *s && spctabnl (*s); s++) + ; + if (s == 0 || *s == 0) + return ((WORD_LIST *)NULL); + + s_len = strlen (s); + tokstart = i = 0; + list = (WORD_LIST *)NULL; + while (1) + { + c = s[i]; + if (c == '\\') + { + i++; + if (s[i]) + i++; + } + else if (c == '\'') + i = skip_single_quoted (s, s_len, ++i); + else if (c == '"') + i = skip_double_quoted (s, s_len, ++i); + else if (c == 0 || spctabnl (c)) + { + /* We have found the end of a token. Make a word out of it and + add it to the word list. */ + token = substring (s, tokstart, i); + list = add_string_to_list (token, list); + free (token); + while (spctabnl (s[i])) + i++; + if (s[i]) + tokstart = i; + else + break; + } + else + i++; /* normal character */ + } + return (REVERSE_LIST (list, WORD_LIST *)); +} +#endif + +/********************************************************/ +/* */ +/* Functions to perform assignment statements */ +/* */ +/********************************************************/ + +#if defined (ARRAY_VARS) +static SHELL_VAR * +do_compound_assignment (name, value, flags) + char *name, *value; + int flags; +{ + SHELL_VAR *v; + int mklocal, mkassoc, mkglobal; + WORD_LIST *list; + + mklocal = flags & ASS_MKLOCAL; + mkassoc = flags & ASS_MKASSOC; + mkglobal = flags & ASS_MKGLOBAL; + + if (mklocal && variable_context) + { + v = find_variable (name); + list = expand_compound_array_assignment (v, value, flags); + if (mkassoc) + v = make_local_assoc_variable (name); + else if (v == 0 || (array_p (v) == 0 && assoc_p (v) == 0) || v->context != variable_context) + v = make_local_array_variable (name, 0); + if (v) + assign_compound_array_list (v, list, flags); + } + /* In a function but forcing assignment in global context */ + else if (mkglobal && variable_context) + { + v = find_global_variable (name); + list = expand_compound_array_assignment (v, value, flags); + if (v == 0 && mkassoc) + v = make_new_assoc_variable (name); + else if (v && mkassoc && assoc_p (v) == 0) + v = convert_var_to_assoc (v); + else if (v == 0) + v = make_new_array_variable (name); + else if (v && mkassoc == 0 && array_p (v) == 0) + v = convert_var_to_array (v); + if (v) + assign_compound_array_list (v, list, flags); + } + else + v = assign_array_from_string (name, value, flags); + + return (v); +} +#endif + +/* Given STRING, an assignment string, get the value of the right side + of the `=', and bind it to the left side. If EXPAND is true, then + perform parameter expansion, command substitution, and arithmetic + expansion on the right-hand side. Perform tilde expansion in any + case. Do not perform word splitting on the result of expansion. */ +static int +do_assignment_internal (word, expand) + const WORD_DESC *word; + int expand; +{ + int offset, appendop, assign_list, aflags, retval; + char *name, *value, *temp; + SHELL_VAR *entry; +#if defined (ARRAY_VARS) + char *t; + int ni; +#endif + const char *string; + + if (word == 0 || word->word == 0) + return 0; + + appendop = assign_list = aflags = 0; + string = word->word; + offset = assignment (string, 0); + name = savestring (string); + value = (char *)NULL; + + if (name[offset] == '=') + { + if (name[offset - 1] == '+') + { + appendop = 1; + name[offset - 1] = '\0'; + } + + name[offset] = 0; /* might need this set later */ + temp = name + offset + 1; + +#if defined (ARRAY_VARS) + if (expand && (word->flags & W_COMPASSIGN)) + { + assign_list = ni = 1; + value = extract_array_assignment_list (temp, &ni); + } + else +#endif + if (expand && temp[0]) + value = expand_string_if_necessary (temp, 0, expand_string_assignment); + else + value = savestring (temp); + } + + if (value == 0) + { + value = (char *)xmalloc (1); + value[0] = '\0'; + } + + if (echo_command_at_execute) + { + if (appendop) + name[offset - 1] = '+'; + xtrace_print_assignment (name, value, assign_list, 1); + if (appendop) + name[offset - 1] = '\0'; + } + +#define ASSIGN_RETURN(r) do { FREE (value); free (name); return (r); } while (0) + + if (appendop) + aflags |= ASS_APPEND; + +#if defined (ARRAY_VARS) + if (t = mbschr (name, '[')) /*]*/ + { + if (assign_list) + { + report_error (_("%s: cannot assign list to array member"), name); + ASSIGN_RETURN (0); + } + entry = assign_array_element (name, value, aflags); + if (entry == 0) + ASSIGN_RETURN (0); + } + else if (assign_list) + { + if ((word->flags & W_ASSIGNARG) && (word->flags & W_ASSNGLOBAL) == 0) + aflags |= ASS_MKLOCAL; + if ((word->flags & W_ASSIGNARG) && (word->flags & W_ASSNGLOBAL)) + aflags |= ASS_MKGLOBAL; + if (word->flags & W_ASSIGNASSOC) + aflags |= ASS_MKASSOC; + entry = do_compound_assignment (name, value, aflags); + } + else +#endif /* ARRAY_VARS */ + entry = bind_variable (name, value, aflags); + + stupidly_hack_special_variables (name); + + /* Return 1 if the assignment seems to have been performed correctly. */ + if (entry == 0 || readonly_p (entry)) + retval = 0; /* assignment failure */ + else if (noassign_p (entry)) + { + last_command_exit_value = EXECUTION_FAILURE; + retval = 1; /* error status, but not assignment failure */ + } + else + retval = 1; + + if (entry && retval != 0 && noassign_p (entry) == 0) + VUNSETATTR (entry, att_invisible); + + ASSIGN_RETURN (retval); +} + +/* Perform the assignment statement in STRING, and expand the + right side by doing tilde, command and parameter expansion. */ +int +do_assignment (string) + char *string; +{ + WORD_DESC td; + + td.flags = W_ASSIGNMENT; + td.word = string; + + return do_assignment_internal (&td, 1); +} + +int +do_word_assignment (word, flags) + WORD_DESC *word; + int flags; +{ + return do_assignment_internal (word, 1); +} + +/* Given STRING, an assignment string, get the value of the right side + of the `=', and bind it to the left side. Do not perform any word + expansions on the right hand side. */ +int +do_assignment_no_expand (string) + char *string; +{ + WORD_DESC td; + + td.flags = W_ASSIGNMENT; + td.word = string; + + return (do_assignment_internal (&td, 0)); +} + +/*************************************************** + * * + * Functions to manage the positional parameters * + * * + ***************************************************/ + +/* Return the word list that corresponds to `$*'. */ +WORD_LIST * +list_rest_of_args () +{ + register WORD_LIST *list, *args; + int i; + + /* Break out of the loop as soon as one of the dollar variables is null. */ + for (i = 1, list = (WORD_LIST *)NULL; i < 10 && dollar_vars[i]; i++) + list = make_word_list (make_bare_word (dollar_vars[i]), list); + + for (args = rest_of_args; args; args = args->next) + list = make_word_list (make_bare_word (args->word->word), list); + + return (REVERSE_LIST (list, WORD_LIST *)); +} + +int +number_of_args () +{ + register WORD_LIST *list; + int n; + + for (n = 0; n < 9 && dollar_vars[n+1]; n++) + ; + for (list = rest_of_args; list; list = list->next) + n++; + return n; +} + +/* Return the value of a positional parameter. This handles values > 10. */ +char * +get_dollar_var_value (ind) + intmax_t ind; +{ + char *temp; + WORD_LIST *p; + + if (ind < 10) + temp = dollar_vars[ind] ? savestring (dollar_vars[ind]) : (char *)NULL; + else /* We want something like ${11} */ + { + ind -= 10; + for (p = rest_of_args; p && ind--; p = p->next) + ; + temp = p ? savestring (p->word->word) : (char *)NULL; + } + return (temp); +} + +/* Make a single large string out of the dollar digit variables, + and the rest_of_args. If DOLLAR_STAR is 1, then obey the special + case of "$*" with respect to IFS. */ +char * +string_rest_of_args (dollar_star) + int dollar_star; +{ + register WORD_LIST *list; + char *string; + + list = list_rest_of_args (); + string = dollar_star ? string_list_dollar_star (list) : string_list (list); + dispose_words (list); + return (string); +} + +/* Return a string containing the positional parameters from START to + END, inclusive. If STRING[0] == '*', we obey the rules for $*, + which only makes a difference if QUOTED is non-zero. If QUOTED includes + Q_HERE_DOCUMENT or Q_DOUBLE_QUOTES, this returns a quoted list, otherwise + no quoting chars are added. */ +static char * +pos_params (string, start, end, quoted) + char *string; + int start, end, quoted; +{ + WORD_LIST *save, *params, *h, *t; + char *ret; + int i; + + /* see if we can short-circuit. if start == end, we want 0 parameters. */ + if (start == end) + return ((char *)NULL); + + save = params = list_rest_of_args (); + if (save == 0 && start > 0) + return ((char *)NULL); + + if (start == 0) /* handle ${@:0[:x]} specially */ + { + t = make_word_list (make_word (dollar_vars[0]), params); + save = params = t; + } + + for (i = start ? 1 : 0; params && i < start; i++) + params = params->next; + if (params == 0) + { + dispose_words (save); + return ((char *)NULL); + } + for (h = t = params; params && i < end; i++) + { + t = params; + params = params->next; + } + + t->next = (WORD_LIST *)NULL; + + ret = string_list_pos_params (string[0], h, quoted); + + if (t != params) + t->next = params; + + dispose_words (save); + return (ret); +} + +/******************************************************************/ +/* */ +/* Functions to expand strings to strings or WORD_LISTs */ +/* */ +/******************************************************************/ + +#if defined (PROCESS_SUBSTITUTION) +#define EXP_CHAR(s) (s == '$' || s == '`' || s == '<' || s == '>' || s == CTLESC || s == '~') +#else +#define EXP_CHAR(s) (s == '$' || s == '`' || s == CTLESC || s == '~') +#endif + +/* If there are any characters in STRING that require full expansion, + then call FUNC to expand STRING; otherwise just perform quote + removal if necessary. This returns a new string. */ +static char * +expand_string_if_necessary (string, quoted, func) + char *string; + int quoted; + EXPFUNC *func; +{ + WORD_LIST *list; + size_t slen; + int i, saw_quote; + char *ret; + DECLARE_MBSTATE; + + /* Don't need string length for ADVANCE_CHAR unless multibyte chars possible. */ + slen = (MB_CUR_MAX > 1) ? strlen (string) : 0; + i = saw_quote = 0; + while (string[i]) + { + if (EXP_CHAR (string[i])) + break; + else if (string[i] == '\'' || string[i] == '\\' || string[i] == '"') + saw_quote = 1; + ADVANCE_CHAR (string, slen, i); + } + + if (string[i]) + { + list = (*func) (string, quoted); + if (list) + { + ret = string_list (list); + dispose_words (list); + } + else + ret = (char *)NULL; + } + else if (saw_quote && ((quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) == 0)) + ret = string_quote_removal (string, quoted); + else + ret = savestring (string); + + return ret; +} + +static inline char * +expand_string_to_string_internal (string, quoted, func) + char *string; + int quoted; + EXPFUNC *func; +{ + WORD_LIST *list; + char *ret; + + if (string == 0 || *string == '\0') + return ((char *)NULL); + + list = (*func) (string, quoted); + if (list) + { + ret = string_list (list); + dispose_words (list); + } + else + ret = (char *)NULL; + + return (ret); +} + +char * +expand_string_to_string (string, quoted) + char *string; + int quoted; +{ + return (expand_string_to_string_internal (string, quoted, expand_string)); +} + +char * +expand_string_unsplit_to_string (string, quoted) + char *string; + int quoted; +{ + return (expand_string_to_string_internal (string, quoted, expand_string_unsplit)); +} + +char * +expand_assignment_string_to_string (string, quoted) + char *string; + int quoted; +{ + return (expand_string_to_string_internal (string, quoted, expand_string_assignment)); +} + +char * +expand_arith_string (string, quoted) + char *string; + int quoted; +{ + WORD_DESC td; + WORD_LIST *list, *tlist; + size_t slen; + int i, saw_quote; + char *ret; + DECLARE_MBSTATE; + + /* Don't need string length for ADVANCE_CHAR unless multibyte chars possible. */ + slen = (MB_CUR_MAX > 1) ? strlen (string) : 0; + i = saw_quote = 0; + while (string[i]) + { + if (EXP_CHAR (string[i])) + break; + else if (string[i] == '\'' || string[i] == '\\' || string[i] == '"') + saw_quote = 1; + ADVANCE_CHAR (string, slen, i); + } + + if (string[i]) + { + /* This is expanded version of expand_string_internal as it's called by + expand_string_leave_quoted */ + td.flags = W_NOPROCSUB; /* don't want process substitution */ + td.word = savestring (string); + list = call_expand_word_internal (&td, quoted, 0, (int *)NULL, (int *)NULL); + /* This takes care of the calls from expand_string_leave_quoted and + expand_string */ + if (list) + { + tlist = word_list_split (list); + dispose_words (list); + list = tlist; + if (list) + dequote_list (list); + } + /* This comes from expand_string_if_necessary */ + if (list) + { + ret = string_list (list); + dispose_words (list); + } + else + ret = (char *)NULL; + FREE (td.word); + } + else if (saw_quote && (quoted & Q_ARITH)) + ret = string_quote_removal (string, quoted); + else if (saw_quote && ((quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) == 0)) + ret = string_quote_removal (string, quoted); + else + ret = savestring (string); + + return ret; +} + +#if defined (COND_COMMAND) +/* Just remove backslashes in STRING. Returns a new string. */ +char * +remove_backslashes (string) + char *string; +{ + char *r, *ret, *s; + + r = ret = (char *)xmalloc (strlen (string) + 1); + for (s = string; s && *s; ) + { + if (*s == '\\') + s++; + if (*s == 0) + break; + *r++ = *s++; + } + *r = '\0'; + return ret; +} + +/* This needs better error handling. */ +/* Expand W for use as an argument to a unary or binary operator in a + [[...]] expression. If SPECIAL is 1, this is the rhs argument + to the != or == operator, and should be treated as a pattern. In + this case, we quote the string specially for the globbing code. If + SPECIAL is 2, this is an rhs argument for the =~ operator, and should + be quoted appropriately for regcomp/regexec. The caller is responsible + for removing the backslashes if the unquoted word is needed later. */ +char * +cond_expand_word (w, special) + WORD_DESC *w; + int special; +{ + char *r, *p; + WORD_LIST *l; + int qflags; + + if (w->word == 0 || w->word[0] == '\0') + return ((char *)NULL); + + expand_no_split_dollar_star = 1; + w->flags |= W_NOSPLIT2; + l = call_expand_word_internal (w, 0, 0, (int *)0, (int *)0); + expand_no_split_dollar_star = 0; + if (l) + { + if (special == 0) /* LHS */ + { + dequote_list (l); + r = string_list (l); + } + else + { + /* Need to figure out whether or not we should call dequote_escapes + or a new dequote_ctlnul function here, and under what + circumstances. */ + qflags = QGLOB_CVTNULL; + if (special == 2) + qflags |= QGLOB_REGEXP; + p = string_list (l); + r = quote_string_for_globbing (p, qflags); + free (p); + } + dispose_words (l); + } + else + r = (char *)NULL; + + return r; +} +#endif + +/* Call expand_word_internal to expand W and handle error returns. + A convenience function for functions that don't want to handle + any errors or free any memory before aborting. */ +static WORD_LIST * +call_expand_word_internal (w, q, i, c, e) + WORD_DESC *w; + int q, i, *c, *e; +{ + WORD_LIST *result; + + result = expand_word_internal (w, q, i, c, e); + if (result == &expand_word_error || result == &expand_word_fatal) + { + /* By convention, each time this error is returned, w->word has + already been freed (it sometimes may not be in the fatal case, + but that doesn't result in a memory leak because we're going + to exit in most cases). */ + w->word = (char *)NULL; + last_command_exit_value = EXECUTION_FAILURE; + exp_jump_to_top_level ((result == &expand_word_error) ? DISCARD : FORCE_EOF); + /* NOTREACHED */ + return (NULL); + } + else + return (result); +} + +/* Perform parameter expansion, command substitution, and arithmetic + expansion on STRING, as if it were a word. Leave the result quoted. + Since this does not perform word splitting, it leaves quoted nulls + in the result. */ +static WORD_LIST * +expand_string_internal (string, quoted) + char *string; + int quoted; +{ + WORD_DESC td; + WORD_LIST *tresult; + + if (string == 0 || *string == 0) + return ((WORD_LIST *)NULL); + + td.flags = 0; + td.word = savestring (string); + + tresult = call_expand_word_internal (&td, quoted, 0, (int *)NULL, (int *)NULL); + + FREE (td.word); + return (tresult); +} + +/* Expand STRING by performing parameter expansion, command substitution, + and arithmetic expansion. Dequote the resulting WORD_LIST before + returning it, but do not perform word splitting. The call to + remove_quoted_nulls () is in here because word splitting normally + takes care of quote removal. */ +WORD_LIST * +expand_string_unsplit (string, quoted) + char *string; + int quoted; +{ + WORD_LIST *value; + + if (string == 0 || *string == '\0') + return ((WORD_LIST *)NULL); + + expand_no_split_dollar_star = 1; + value = expand_string_internal (string, quoted); + expand_no_split_dollar_star = 0; + + if (value) + { + if (value->word) + { + remove_quoted_nulls (value->word->word); + value->word->flags &= ~W_HASQUOTEDNULL; + } + dequote_list (value); + } + return (value); +} + +/* Expand the rhs of an assignment statement */ +WORD_LIST * +expand_string_assignment (string, quoted) + char *string; + int quoted; +{ + WORD_DESC td; + WORD_LIST *value; + + if (string == 0 || *string == '\0') + return ((WORD_LIST *)NULL); + + expand_no_split_dollar_star = 1; + + td.flags = W_ASSIGNRHS; + td.word = savestring (string); + value = call_expand_word_internal (&td, quoted, 0, (int *)NULL, (int *)NULL); + FREE (td.word); + + expand_no_split_dollar_star = 0; + + if (value) + { + if (value->word) + { + remove_quoted_nulls (value->word->word); + value->word->flags &= ~W_HASQUOTEDNULL; + } + dequote_list (value); + } + return (value); +} + + +/* Expand one of the PS? prompt strings. This is a sort of combination of + expand_string_unsplit and expand_string_internal, but returns the + passed string when an error occurs. Might want to trap other calls + to jump_to_top_level here so we don't endlessly loop. */ +WORD_LIST * +expand_prompt_string (string, quoted, wflags) + char *string; + int quoted; + int wflags; +{ + WORD_LIST *value; + WORD_DESC td; + + if (string == 0 || *string == 0) + return ((WORD_LIST *)NULL); + + td.flags = wflags; + td.word = savestring (string); + + no_longjmp_on_fatal_error = 1; + value = expand_word_internal (&td, quoted, 0, (int *)NULL, (int *)NULL); + no_longjmp_on_fatal_error = 0; + + if (value == &expand_word_error || value == &expand_word_fatal) + { + value = make_word_list (make_bare_word (string), (WORD_LIST *)NULL); + return value; + } + FREE (td.word); + if (value) + { + if (value->word) + { + remove_quoted_nulls (value->word->word); + value->word->flags &= ~W_HASQUOTEDNULL; + } + dequote_list (value); + } + return (value); +} + +/* Expand STRING just as if you were expanding a word, but do not dequote + the resultant WORD_LIST. This is called only from within this file, + and is used to correctly preserve quoted characters when expanding + things like ${1+"$@"}. This does parameter expansion, command + substitution, arithmetic expansion, and word splitting. */ +static WORD_LIST * +expand_string_leave_quoted (string, quoted) + char *string; + int quoted; +{ + WORD_LIST *tlist; + WORD_LIST *tresult; + + if (string == 0 || *string == '\0') + return ((WORD_LIST *)NULL); + + tlist = expand_string_internal (string, quoted); + + if (tlist) + { + tresult = word_list_split (tlist); + dispose_words (tlist); + return (tresult); + } + return ((WORD_LIST *)NULL); +} + +/* This does not perform word splitting or dequote the WORD_LIST + it returns. */ +static WORD_LIST * +expand_string_for_rhs (string, quoted, dollar_at_p, has_dollar_at) + char *string; + int quoted, *dollar_at_p, *has_dollar_at; +{ + WORD_DESC td; + WORD_LIST *tresult; + + if (string == 0 || *string == '\0') + return (WORD_LIST *)NULL; + + expand_no_split_dollar_star = 1; + td.flags = W_NOSPLIT2; /* no splitting, remove "" and '' */ + td.word = string; + tresult = call_expand_word_internal (&td, quoted, 1, dollar_at_p, has_dollar_at); + expand_no_split_dollar_star = 0; + + return (tresult); +} + +/* Expand STRING just as if you were expanding a word. This also returns + a list of words. Note that filename globbing is *NOT* done for word + or string expansion, just when the shell is expanding a command. This + does parameter expansion, command substitution, arithmetic expansion, + and word splitting. Dequote the resultant WORD_LIST before returning. */ +WORD_LIST * +expand_string (string, quoted) + char *string; + int quoted; +{ + WORD_LIST *result; + + if (string == 0 || *string == '\0') + return ((WORD_LIST *)NULL); + + result = expand_string_leave_quoted (string, quoted); + return (result ? dequote_list (result) : result); +} + +/*************************************************** + * * + * Functions to handle quoting chars * + * * + ***************************************************/ + +/* Conventions: + + A string with s[0] == CTLNUL && s[1] == 0 is a quoted null string. + The parser passes CTLNUL as CTLESC CTLNUL. */ + +/* Quote escape characters in string s, but no other characters. This is + used to protect CTLESC and CTLNUL in variable values from the rest of + the word expansion process after the variable is expanded (word splitting + and filename generation). If IFS is null, we quote spaces as well, just + in case we split on spaces later (in the case of unquoted $@, we will + eventually attempt to split the entire word on spaces). Corresponding + code exists in dequote_escapes. Even if we don't end up splitting on + spaces, quoting spaces is not a problem. This should never be called on + a string that is quoted with single or double quotes or part of a here + document (effectively double-quoted). */ +char * +quote_escapes (string) + char *string; +{ + register char *s, *t; + size_t slen; + char *result, *send; + int quote_spaces, skip_ctlesc, skip_ctlnul; + DECLARE_MBSTATE; + + slen = strlen (string); + send = string + slen; + + quote_spaces = (ifs_value && *ifs_value == 0); + + for (skip_ctlesc = skip_ctlnul = 0, s = ifs_value; s && *s; s++) + skip_ctlesc |= *s == CTLESC, skip_ctlnul |= *s == CTLNUL; + + t = result = (char *)xmalloc ((slen * 2) + 1); + s = string; + + while (*s) + { + if ((skip_ctlesc == 0 && *s == CTLESC) || (skip_ctlnul == 0 && *s == CTLNUL) || (quote_spaces && *s == ' ')) + *t++ = CTLESC; + COPY_CHAR_P (t, s, send); + } + *t = '\0'; + + return (result); +} + +static WORD_LIST * +list_quote_escapes (list) + WORD_LIST *list; +{ + register WORD_LIST *w; + char *t; + + for (w = list; w; w = w->next) + { + t = w->word->word; + w->word->word = quote_escapes (t); + free (t); + } + return list; +} + +/* Inverse of quote_escapes; remove CTLESC protecting CTLESC or CTLNUL. + + The parser passes us CTLESC as CTLESC CTLESC and CTLNUL as CTLESC CTLNUL. + This is necessary to make unquoted CTLESC and CTLNUL characters in the + data stream pass through properly. + + We need to remove doubled CTLESC characters inside quoted strings before + quoting the entire string, so we do not double the number of CTLESC + characters. + + Also used by parts of the pattern substitution code. */ +char * +dequote_escapes (string) + char *string; +{ + register char *s, *t, *s1; + size_t slen; + char *result, *send; + int quote_spaces; + DECLARE_MBSTATE; + + if (string == 0) + return string; + + slen = strlen (string); + send = string + slen; + + t = result = (char *)xmalloc (slen + 1); + + if (strchr (string, CTLESC) == 0) + return (strcpy (result, string)); + + quote_spaces = (ifs_value && *ifs_value == 0); + + s = string; + while (*s) + { + if (*s == CTLESC && (s[1] == CTLESC || s[1] == CTLNUL || (quote_spaces && s[1] == ' '))) + { + s++; + if (*s == '\0') + break; + } + COPY_CHAR_P (t, s, send); + } + *t = '\0'; + + return result; +} + +static WORD_LIST * +list_dequote_escapes (list) + WORD_LIST *list; +{ + register WORD_LIST *w; + char *t; + + for (w = list; w; w = w->next) + { + t = w->word->word; + w->word->word = dequote_escapes (t); + free (t); + } + return list; +} + +/* Return a new string with the quoted representation of character C. + This turns "" into QUOTED_NULL, so the W_HASQUOTEDNULL flag needs to be + set in any resultant WORD_DESC where this value is the word. */ +static char * +make_quoted_char (c) + int c; +{ + char *temp; + + temp = (char *)xmalloc (3); + if (c == 0) + { + temp[0] = CTLNUL; + temp[1] = '\0'; + } + else + { + temp[0] = CTLESC; + temp[1] = c; + temp[2] = '\0'; + } + return (temp); +} + +/* Quote STRING, returning a new string. This turns "" into QUOTED_NULL, so + the W_HASQUOTEDNULL flag needs to be set in any resultant WORD_DESC where + this value is the word. */ +char * +quote_string (string) + char *string; +{ + register char *t; + size_t slen; + char *result, *send; + + if (*string == 0) + { + result = (char *)xmalloc (2); + result[0] = CTLNUL; + result[1] = '\0'; + } + else + { + DECLARE_MBSTATE; + + slen = strlen (string); + send = string + slen; + + result = (char *)xmalloc ((slen * 2) + 1); + + for (t = result; string < send; ) + { + *t++ = CTLESC; + COPY_CHAR_P (t, string, send); + } + *t = '\0'; + } + return (result); +} + +/* De-quote quoted characters in STRING. */ +char * +dequote_string (string) + char *string; +{ + register char *s, *t; + size_t slen; + char *result, *send; + DECLARE_MBSTATE; + + slen = strlen (string); + + t = result = (char *)xmalloc (slen + 1); + + if (QUOTED_NULL (string)) + { + result[0] = '\0'; + return (result); + } + + /* If no character in the string can be quoted, don't bother examining + each character. Just return a copy of the string passed to us. */ + if (strchr (string, CTLESC) == NULL) + return (strcpy (result, string)); + + send = string + slen; + s = string; + while (*s) + { + if (*s == CTLESC) + { + s++; + if (*s == '\0') + break; + } + COPY_CHAR_P (t, s, send); + } + + *t = '\0'; + return (result); +} + +/* Quote the entire WORD_LIST list. */ +static WORD_LIST * +quote_list (list) + WORD_LIST *list; +{ + register WORD_LIST *w; + char *t; + + for (w = list; w; w = w->next) + { + t = w->word->word; + w->word->word = quote_string (t); + if (*t == 0) + w->word->flags |= W_HASQUOTEDNULL; /* XXX - turn on W_HASQUOTEDNULL here? */ + w->word->flags |= W_QUOTED; + free (t); + } + return list; +} + +/* De-quote quoted characters in each word in LIST. */ +WORD_LIST * +dequote_list (list) + WORD_LIST *list; +{ + register char *s; + register WORD_LIST *tlist; + + for (tlist = list; tlist; tlist = tlist->next) + { + s = dequote_string (tlist->word->word); + if (QUOTED_NULL (tlist->word->word)) + tlist->word->flags &= ~W_HASQUOTEDNULL; + free (tlist->word->word); + tlist->word->word = s; + } + return list; +} + +/* Remove CTLESC protecting a CTLESC or CTLNUL in place. Return the passed + string. */ +char * +remove_quoted_escapes (string) + char *string; +{ + char *t; + + if (string) + { + t = dequote_escapes (string); + strcpy (string, t); + free (t); + } + + return (string); +} + +/* Perform quoted null character removal on STRING. We don't allow any + quoted null characters in the middle or at the ends of strings because + of how expand_word_internal works. remove_quoted_nulls () turns + STRING into an empty string iff it only consists of a quoted null, + and removes all unquoted CTLNUL characters. */ +char * +remove_quoted_nulls (string) + char *string; +{ + register size_t slen; + register int i, j, prev_i; + DECLARE_MBSTATE; + + if (strchr (string, CTLNUL) == 0) /* XXX */ + return string; /* XXX */ + + slen = strlen (string); + i = j = 0; + + while (i < slen) + { + if (string[i] == CTLESC) + { + /* Old code had j++, but we cannot assume that i == j at this + point -- what if a CTLNUL has already been removed from the + string? We don't want to drop the CTLESC or recopy characters + that we've already copied down. */ + i++; string[j++] = CTLESC; + if (i == slen) + break; + } + else if (string[i] == CTLNUL) + { + i++; + continue; + } + + prev_i = i; + ADVANCE_CHAR (string, slen, i); + if (j < prev_i) + { + do string[j++] = string[prev_i++]; while (prev_i < i); + } + else + j = i; + } + string[j] = '\0'; + + return (string); +} + +/* Perform quoted null character removal on each element of LIST. + This modifies LIST. */ +void +word_list_remove_quoted_nulls (list) + WORD_LIST *list; +{ + register WORD_LIST *t; + + for (t = list; t; t = t->next) + { + remove_quoted_nulls (t->word->word); + t->word->flags &= ~W_HASQUOTEDNULL; + } +} + +/* **************************************************************** */ +/* */ +/* Functions for Matching and Removing Patterns */ +/* */ +/* **************************************************************** */ + +#if defined (HANDLE_MULTIBYTE) +#if 0 /* Currently unused */ +static unsigned char * +mb_getcharlens (string, len) + char *string; + int len; +{ + int i, offset, last; + unsigned char *ret; + char *p; + DECLARE_MBSTATE; + + i = offset = 0; + last = 0; + ret = (unsigned char *)xmalloc (len); + memset (ret, 0, len); + while (string[last]) + { + ADVANCE_CHAR (string, len, offset); + ret[last] = offset - last; + last = offset; + } + return ret; +} +#endif +#endif + +/* Remove the portion of PARAM matched by PATTERN according to OP, where OP + can have one of 4 values: + RP_LONG_LEFT remove longest matching portion at start of PARAM + RP_SHORT_LEFT remove shortest matching portion at start of PARAM + RP_LONG_RIGHT remove longest matching portion at end of PARAM + RP_SHORT_RIGHT remove shortest matching portion at end of PARAM +*/ + +#define RP_LONG_LEFT 1 +#define RP_SHORT_LEFT 2 +#define RP_LONG_RIGHT 3 +#define RP_SHORT_RIGHT 4 + +/* Returns its first argument if nothing matched; new memory otherwise */ +static char * +remove_upattern (param, pattern, op) + char *param, *pattern; + int op; +{ + register int len; + register char *end; + register char *p, *ret, c; + + len = STRLEN (param); + end = param + len; + + switch (op) + { + case RP_LONG_LEFT: /* remove longest match at start */ + for (p = end; p >= param; p--) + { + c = *p; *p = '\0'; + if (strmatch (pattern, param, FNMATCH_EXTFLAG) != FNM_NOMATCH) + { + *p = c; + return (savestring (p)); + } + *p = c; + + } + break; + + case RP_SHORT_LEFT: /* remove shortest match at start */ + for (p = param; p <= end; p++) + { + c = *p; *p = '\0'; + if (strmatch (pattern, param, FNMATCH_EXTFLAG) != FNM_NOMATCH) + { + *p = c; + return (savestring (p)); + } + *p = c; + } + break; + + case RP_LONG_RIGHT: /* remove longest match at end */ + for (p = param; p <= end; p++) + { + if (strmatch (pattern, p, FNMATCH_EXTFLAG) != FNM_NOMATCH) + { + c = *p; *p = '\0'; + ret = savestring (param); + *p = c; + return (ret); + } + } + break; + + case RP_SHORT_RIGHT: /* remove shortest match at end */ + for (p = end; p >= param; p--) + { + if (strmatch (pattern, p, FNMATCH_EXTFLAG) != FNM_NOMATCH) + { + c = *p; *p = '\0'; + ret = savestring (param); + *p = c; + return (ret); + } + } + break; + } + + return (param); /* no match, return original string */ +} + +#if defined (HANDLE_MULTIBYTE) +/* Returns its first argument if nothing matched; new memory otherwise */ +static wchar_t * +remove_wpattern (wparam, wstrlen, wpattern, op) + wchar_t *wparam; + size_t wstrlen; + wchar_t *wpattern; + int op; +{ + wchar_t wc, *ret; + int n; + + switch (op) + { + case RP_LONG_LEFT: /* remove longest match at start */ + for (n = wstrlen; n >= 0; n--) + { + wc = wparam[n]; wparam[n] = L'\0'; + if (wcsmatch (wpattern, wparam, FNMATCH_EXTFLAG) != FNM_NOMATCH) + { + wparam[n] = wc; + return (wcsdup (wparam + n)); + } + wparam[n] = wc; + } + break; + + case RP_SHORT_LEFT: /* remove shortest match at start */ + for (n = 0; n <= wstrlen; n++) + { + wc = wparam[n]; wparam[n] = L'\0'; + if (wcsmatch (wpattern, wparam, FNMATCH_EXTFLAG) != FNM_NOMATCH) + { + wparam[n] = wc; + return (wcsdup (wparam + n)); + } + wparam[n] = wc; + } + break; + + case RP_LONG_RIGHT: /* remove longest match at end */ + for (n = 0; n <= wstrlen; n++) + { + if (wcsmatch (wpattern, wparam + n, FNMATCH_EXTFLAG) != FNM_NOMATCH) + { + wc = wparam[n]; wparam[n] = L'\0'; + ret = wcsdup (wparam); + wparam[n] = wc; + return (ret); + } + } + break; + + case RP_SHORT_RIGHT: /* remove shortest match at end */ + for (n = wstrlen; n >= 0; n--) + { + if (wcsmatch (wpattern, wparam + n, FNMATCH_EXTFLAG) != FNM_NOMATCH) + { + wc = wparam[n]; wparam[n] = L'\0'; + ret = wcsdup (wparam); + wparam[n] = wc; + return (ret); + } + } + break; + } + + return (wparam); /* no match, return original string */ +} +#endif /* HANDLE_MULTIBYTE */ + +static char * +remove_pattern (param, pattern, op) + char *param, *pattern; + int op; +{ + char *xret; + + if (param == NULL) + return (param); + if (*param == '\0' || pattern == NULL || *pattern == '\0') /* minor optimization */ + return (savestring (param)); + +#if defined (HANDLE_MULTIBYTE) + if (MB_CUR_MAX > 1) + { + wchar_t *ret, *oret; + size_t n; + wchar_t *wparam, *wpattern; + mbstate_t ps; + + n = xdupmbstowcs (&wpattern, NULL, pattern); + if (n == (size_t)-1) + { + xret = remove_upattern (param, pattern, op); + return ((xret == param) ? savestring (param) : xret); + } + n = xdupmbstowcs (&wparam, NULL, param); + + if (n == (size_t)-1) + { + free (wpattern); + xret = remove_upattern (param, pattern, op); + return ((xret == param) ? savestring (param) : xret); + } + oret = ret = remove_wpattern (wparam, n, wpattern, op); + /* Don't bother to convert wparam back to multibyte string if nothing + matched; just return copy of original string */ + if (ret == wparam) + { + free (wparam); + free (wpattern); + return (savestring (param)); + } + + free (wparam); + free (wpattern); + + n = strlen (param); + xret = (char *)xmalloc (n + 1); + memset (&ps, '\0', sizeof (mbstate_t)); + n = wcsrtombs (xret, (const wchar_t **)&ret, n, &ps); + xret[n] = '\0'; /* just to make sure */ + free (oret); + return xret; + } + else +#endif + { + xret = remove_upattern (param, pattern, op); + return ((xret == param) ? savestring (param) : xret); + } +} + +/* Match PAT anywhere in STRING and return the match boundaries. + This returns 1 in case of a successful match, 0 otherwise. SP + and EP are pointers into the string where the match begins and + ends, respectively. MTYPE controls what kind of match is attempted. + MATCH_BEG and MATCH_END anchor the match at the beginning and end + of the string, respectively. The longest match is returned. */ +static int +match_upattern (string, pat, mtype, sp, ep) + char *string, *pat; + int mtype; + char **sp, **ep; +{ + int c, len, mlen; + register char *p, *p1, *npat; + char *end; + int n1; + + /* If the pattern doesn't match anywhere in the string, go ahead and + short-circuit right away. A minor optimization, saves a bunch of + unnecessary calls to strmatch (up to N calls for a string of N + characters) if the match is unsuccessful. To preserve the semantics + of the substring matches below, we make sure that the pattern has + `*' as first and last character, making a new pattern if necessary. */ + /* XXX - check this later if I ever implement `**' with special meaning, + since this will potentially result in `**' at the beginning or end */ + len = STRLEN (pat); + if (pat[0] != '*' || (pat[0] == '*' && pat[1] == LPAREN && extended_glob) || pat[len - 1] != '*') + { + p = npat = (char *)xmalloc (len + 3); + p1 = pat; + if (*p1 != '*' || (*p1 == '*' && p1[1] == LPAREN && extended_glob)) + *p++ = '*'; + while (*p1) + *p++ = *p1++; + if (p1[-1] != '*' || p[-2] == '\\') + *p++ = '*'; + *p = '\0'; + } + else + npat = pat; + c = strmatch (npat, string, FNMATCH_EXTFLAG); + if (npat != pat) + free (npat); + if (c == FNM_NOMATCH) + return (0); + + len = STRLEN (string); + end = string + len; + + mlen = umatchlen (pat, len); + + switch (mtype) + { + case MATCH_ANY: + for (p = string; p <= end; p++) + { + if (match_pattern_char (pat, p)) + { + p1 = (mlen == -1) ? end : p + mlen; + /* p1 - p = length of portion of string to be considered + p = current position in string + mlen = number of characters consumed by match (-1 for entire string) + end = end of string + we want to break immediately if the potential match len + is greater than the number of characters remaining in the + string + */ + if (p1 > end) + break; + for ( ; p1 >= p; p1--) + { + c = *p1; *p1 = '\0'; + if (strmatch (pat, p, FNMATCH_EXTFLAG) == 0) + { + *p1 = c; + *sp = p; + *ep = p1; + return 1; + } + *p1 = c; +#if 1 + /* If MLEN != -1, we have a fixed length pattern. */ + if (mlen != -1) + break; +#endif + } + } + } + + return (0); + + case MATCH_BEG: + if (match_pattern_char (pat, string) == 0) + return (0); + + for (p = (mlen == -1) ? end : string + mlen; p >= string; p--) + { + c = *p; *p = '\0'; + if (strmatch (pat, string, FNMATCH_EXTFLAG) == 0) + { + *p = c; + *sp = string; + *ep = p; + return 1; + } + *p = c; + /* If MLEN != -1, we have a fixed length pattern. */ + if (mlen != -1) + break; + } + + return (0); + + case MATCH_END: + for (p = end - ((mlen == -1) ? len : mlen); p <= end; p++) + { + if (strmatch (pat, p, FNMATCH_EXTFLAG) == 0) + { + *sp = p; + *ep = end; + return 1; + } + /* If MLEN != -1, we have a fixed length pattern. */ + if (mlen != -1) + break; + } + + return (0); + } + + return (0); +} + +#if defined (HANDLE_MULTIBYTE) +/* Match WPAT anywhere in WSTRING and return the match boundaries. + This returns 1 in case of a successful match, 0 otherwise. Wide + character version. */ +static int +match_wpattern (wstring, indices, wstrlen, wpat, mtype, sp, ep) + wchar_t *wstring; + char **indices; + size_t wstrlen; + wchar_t *wpat; + int mtype; + char **sp, **ep; +{ + wchar_t wc, *wp, *nwpat, *wp1; + size_t len; + int mlen; + int n, n1, n2, simple; + + simple = (wpat[0] != L'\\' && wpat[0] != L'*' && wpat[0] != L'?' && wpat[0] != L'['); +#if defined (EXTENDED_GLOB) + if (extended_glob) + simple &= (wpat[1] != L'(' || (wpat[0] != L'*' && wpat[0] != L'?' && wpat[0] != L'+' && wpat[0] != L'!' && wpat[0] != L'@')); /*)*/ +#endif + + /* If the pattern doesn't match anywhere in the string, go ahead and + short-circuit right away. A minor optimization, saves a bunch of + unnecessary calls to strmatch (up to N calls for a string of N + characters) if the match is unsuccessful. To preserve the semantics + of the substring matches below, we make sure that the pattern has + `*' as first and last character, making a new pattern if necessary. */ + len = wcslen (wpat); + if (wpat[0] != L'*' || (wpat[0] == L'*' && wpat[1] == WLPAREN && extended_glob) || wpat[len - 1] != L'*') + { + wp = nwpat = (wchar_t *)xmalloc ((len + 3) * sizeof (wchar_t)); + wp1 = wpat; + if (*wp1 != L'*' || (*wp1 == '*' && wp1[1] == WLPAREN && extended_glob)) + *wp++ = L'*'; + while (*wp1 != L'\0') + *wp++ = *wp1++; + if (wp1[-1] != L'*' || wp1[-2] == L'\\') + *wp++ = L'*'; + *wp = '\0'; + } + else + nwpat = wpat; + len = wcsmatch (nwpat, wstring, FNMATCH_EXTFLAG); + if (nwpat != wpat) + free (nwpat); + if (len == FNM_NOMATCH) + return (0); + + mlen = wmatchlen (wpat, wstrlen); + +/* itrace("wmatchlen (%ls) -> %d", wpat, mlen); */ + switch (mtype) + { + case MATCH_ANY: + for (n = 0; n <= wstrlen; n++) + { + n2 = simple ? (*wpat == wstring[n]) : match_pattern_wchar (wpat, wstring + n); + if (n2) + { + n1 = (mlen == -1) ? wstrlen : n + mlen; + if (n1 > wstrlen) + break; + + for ( ; n1 >= n; n1--) + { + wc = wstring[n1]; wstring[n1] = L'\0'; + if (wcsmatch (wpat, wstring + n, FNMATCH_EXTFLAG) == 0) + { + wstring[n1] = wc; + *sp = indices[n]; + *ep = indices[n1]; + return 1; + } + wstring[n1] = wc; + /* If MLEN != -1, we have a fixed length pattern. */ + if (mlen != -1) + break; + } + } + } + + return (0); + + case MATCH_BEG: + if (match_pattern_wchar (wpat, wstring) == 0) + return (0); + + for (n = (mlen == -1) ? wstrlen : mlen; n >= 0; n--) + { + wc = wstring[n]; wstring[n] = L'\0'; + if (wcsmatch (wpat, wstring, FNMATCH_EXTFLAG) == 0) + { + wstring[n] = wc; + *sp = indices[0]; + *ep = indices[n]; + return 1; + } + wstring[n] = wc; + /* If MLEN != -1, we have a fixed length pattern. */ + if (mlen != -1) + break; + } + + return (0); + + case MATCH_END: + for (n = wstrlen - ((mlen == -1) ? wstrlen : mlen); n <= wstrlen; n++) + { + if (wcsmatch (wpat, wstring + n, FNMATCH_EXTFLAG) == 0) + { + *sp = indices[n]; + *ep = indices[wstrlen]; + return 1; + } + /* If MLEN != -1, we have a fixed length pattern. */ + if (mlen != -1) + break; + } + + return (0); + } + + return (0); +} +#endif /* HANDLE_MULTIBYTE */ + +static int +match_pattern (string, pat, mtype, sp, ep) + char *string, *pat; + int mtype; + char **sp, **ep; +{ +#if defined (HANDLE_MULTIBYTE) + int ret; + size_t n; + wchar_t *wstring, *wpat; + char **indices; + size_t slen, plen, mslen, mplen; +#endif + + if (string == 0 || *string == 0 || pat == 0 || *pat == 0) + return (0); + +#if defined (HANDLE_MULTIBYTE) + if (MB_CUR_MAX > 1) + { + if (mbsmbchar (string) == 0 && mbsmbchar (pat) == 0) + return (match_upattern (string, pat, mtype, sp, ep)); + + n = xdupmbstowcs (&wpat, NULL, pat); + if (n == (size_t)-1) + return (match_upattern (string, pat, mtype, sp, ep)); + n = xdupmbstowcs (&wstring, &indices, string); + if (n == (size_t)-1) + { + free (wpat); + return (match_upattern (string, pat, mtype, sp, ep)); + } + ret = match_wpattern (wstring, indices, n, wpat, mtype, sp, ep); + + free (wpat); + free (wstring); + free (indices); + + return (ret); + } + else +#endif + return (match_upattern (string, pat, mtype, sp, ep)); +} + +static int +getpatspec (c, value) + int c; + char *value; +{ + if (c == '#') + return ((*value == '#') ? RP_LONG_LEFT : RP_SHORT_LEFT); + else /* c == '%' */ + return ((*value == '%') ? RP_LONG_RIGHT : RP_SHORT_RIGHT); +} + +/* Posix.2 says that the WORD should be run through tilde expansion, + parameter expansion, command substitution and arithmetic expansion. + This leaves the result quoted, so quote_string_for_globbing () has + to be called to fix it up for strmatch (). If QUOTED is non-zero, + it means that the entire expression was enclosed in double quotes. + This means that quoting characters in the pattern do not make any + special pattern characters quoted. For example, the `*' in the + following retains its special meaning: "${foo#'*'}". */ +static char * +getpattern (value, quoted, expandpat) + char *value; + int quoted, expandpat; +{ + char *pat, *tword; + WORD_LIST *l; +#if 0 + int i; +#endif + /* There is a problem here: how to handle single or double quotes in the + pattern string when the whole expression is between double quotes? + POSIX.2 says that enclosing double quotes do not cause the pattern to + be quoted, but does that leave us a problem with @ and array[@] and their + expansions inside a pattern? */ +#if 0 + if (expandpat && (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) && *tword) + { + i = 0; + pat = string_extract_double_quoted (tword, &i, 1); + free (tword); + tword = pat; + } +#endif + + /* expand_string_for_rhs () leaves WORD quoted and does not perform + word splitting. */ + l = *value ? expand_string_for_rhs (value, + (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) ? Q_PATQUOTE : quoted, + (int *)NULL, (int *)NULL) + : (WORD_LIST *)0; + pat = string_list (l); + dispose_words (l); + if (pat) + { + tword = quote_string_for_globbing (pat, QGLOB_CVTNULL); + free (pat); + pat = tword; + } + return (pat); +} + +#if 0 +/* Handle removing a pattern from a string as a result of ${name%[%]value} + or ${name#[#]value}. */ +static char * +variable_remove_pattern (value, pattern, patspec, quoted) + char *value, *pattern; + int patspec, quoted; +{ + char *tword; + + tword = remove_pattern (value, pattern, patspec); + + return (tword); +} +#endif + +static char * +list_remove_pattern (list, pattern, patspec, itype, quoted) + WORD_LIST *list; + char *pattern; + int patspec, itype, quoted; +{ + WORD_LIST *new, *l; + WORD_DESC *w; + char *tword; + + for (new = (WORD_LIST *)NULL, l = list; l; l = l->next) + { + tword = remove_pattern (l->word->word, pattern, patspec); + w = alloc_word_desc (); + w->word = tword ? tword : savestring (""); + new = make_word_list (w, new); + } + + l = REVERSE_LIST (new, WORD_LIST *); + tword = string_list_pos_params (itype, l, quoted); + dispose_words (l); + + return (tword); +} + +static char * +parameter_list_remove_pattern (itype, pattern, patspec, quoted) + int itype; + char *pattern; + int patspec, quoted; +{ + char *ret; + WORD_LIST *list; + + list = list_rest_of_args (); + if (list == 0) + return ((char *)NULL); + ret = list_remove_pattern (list, pattern, patspec, itype, quoted); + dispose_words (list); + return (ret); +} + +#if defined (ARRAY_VARS) +static char * +array_remove_pattern (var, pattern, patspec, varname, quoted) + SHELL_VAR *var; + char *pattern; + int patspec; + char *varname; /* so we can figure out how it's indexed */ + int quoted; +{ + ARRAY *a; + HASH_TABLE *h; + int itype; + char *ret; + WORD_LIST *list; + SHELL_VAR *v; + + /* compute itype from varname here */ + v = array_variable_part (varname, &ret, 0); + + /* XXX */ + if (v && invisible_p (var)) + return ((char *)NULL); + + itype = ret[0]; + + a = (v && array_p (v)) ? array_cell (v) : 0; + h = (v && assoc_p (v)) ? assoc_cell (v) : 0; + + list = a ? array_to_word_list (a) : (h ? assoc_to_word_list (h) : 0); + if (list == 0) + return ((char *)NULL); + ret = list_remove_pattern (list, pattern, patspec, itype, quoted); + dispose_words (list); + + return ret; +} +#endif /* ARRAY_VARS */ + +static char * +parameter_brace_remove_pattern (varname, value, ind, patstr, rtype, quoted, flags) + char *varname, *value; + int ind; + char *patstr; + int rtype, quoted, flags; +{ + int vtype, patspec, starsub; + char *temp1, *val, *pattern; + SHELL_VAR *v; + + if (value == 0) + return ((char *)NULL); + + this_command_name = varname; + + vtype = get_var_and_type (varname, value, ind, quoted, flags, &v, &val); + if (vtype == -1) + return ((char *)NULL); + + starsub = vtype & VT_STARSUB; + vtype &= ~VT_STARSUB; + + patspec = getpatspec (rtype, patstr); + if (patspec == RP_LONG_LEFT || patspec == RP_LONG_RIGHT) + patstr++; + + /* Need to pass getpattern newly-allocated memory in case of expansion -- + the expansion code will free the passed string on an error. */ + temp1 = savestring (patstr); + pattern = getpattern (temp1, quoted, 1); + free (temp1); + + temp1 = (char *)NULL; /* shut up gcc */ + switch (vtype) + { + case VT_VARIABLE: + case VT_ARRAYMEMBER: + temp1 = remove_pattern (val, pattern, patspec); + if (vtype == VT_VARIABLE) + FREE (val); + if (temp1) + { + val = (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) + ? quote_string (temp1) + : quote_escapes (temp1); + free (temp1); + temp1 = val; + } + break; +#if defined (ARRAY_VARS) + case VT_ARRAYVAR: + temp1 = array_remove_pattern (v, pattern, patspec, varname, quoted); + if (temp1 && ((quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) == 0)) + { + val = quote_escapes (temp1); + free (temp1); + temp1 = val; + } + break; +#endif + case VT_POSPARMS: + temp1 = parameter_list_remove_pattern (varname[0], pattern, patspec, quoted); + if (temp1 && ((quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) == 0)) + { + val = quote_escapes (temp1); + free (temp1); + temp1 = val; + } + break; + } + + FREE (pattern); + return temp1; +} + +/******************************************* + * * + * Functions to expand WORD_DESCs * + * * + *******************************************/ + +/* Expand WORD, performing word splitting on the result. This does + parameter expansion, command substitution, arithmetic expansion, + word splitting, and quote removal. */ + +WORD_LIST * +expand_word (word, quoted) + WORD_DESC *word; + int quoted; +{ + WORD_LIST *result, *tresult; + + tresult = call_expand_word_internal (word, quoted, 0, (int *)NULL, (int *)NULL); + result = word_list_split (tresult); + dispose_words (tresult); + return (result ? dequote_list (result) : result); +} + +/* Expand WORD, but do not perform word splitting on the result. This + does parameter expansion, command substitution, arithmetic expansion, + and quote removal. */ +WORD_LIST * +expand_word_unsplit (word, quoted) + WORD_DESC *word; + int quoted; +{ + WORD_LIST *result; + + expand_no_split_dollar_star = 1; +#if defined (HANDLE_MULTIBYTE) + if (ifs_firstc[0] == 0) +#else + if (ifs_firstc == 0) +#endif + word->flags |= W_NOSPLIT; + word->flags |= W_NOSPLIT2; + result = call_expand_word_internal (word, quoted, 0, (int *)NULL, (int *)NULL); + expand_no_split_dollar_star = 0; + + return (result ? dequote_list (result) : result); +} + +/* Perform shell expansions on WORD, but do not perform word splitting or + quote removal on the result. Virtually identical to expand_word_unsplit; + could be combined if implementations don't diverge. */ +WORD_LIST * +expand_word_leave_quoted (word, quoted) + WORD_DESC *word; + int quoted; +{ + WORD_LIST *result; + + expand_no_split_dollar_star = 1; +#if defined (HANDLE_MULTIBYTE) + if (ifs_firstc[0] == 0) +#else + if (ifs_firstc == 0) +#endif + word->flags |= W_NOSPLIT; + word->flags |= W_NOSPLIT2; + result = call_expand_word_internal (word, quoted, 0, (int *)NULL, (int *)NULL); + expand_no_split_dollar_star = 0; + + return result; +} + +#if defined (PROCESS_SUBSTITUTION) + +/*****************************************************************/ +/* */ +/* Hacking Process Substitution */ +/* */ +/*****************************************************************/ + +#if !defined (HAVE_DEV_FD) +/* Named pipes must be removed explicitly with `unlink'. This keeps a list + of FIFOs the shell has open. unlink_fifo_list will walk the list and + unlink all of them. add_fifo_list adds the name of an open FIFO to the + list. NFIFO is a count of the number of FIFOs in the list. */ +#define FIFO_INCR 20 + +struct temp_fifo { + char *file; + pid_t proc; +}; + +static struct temp_fifo *fifo_list = (struct temp_fifo *)NULL; +static int nfifo; +static int fifo_list_size; + +void +clear_fifo_list () +{ +} + +char * +copy_fifo_list (sizep) + int *sizep; +{ + if (sizep) + *sizep = 0; + return (char *)NULL; +} + +static void +add_fifo_list (pathname) + char *pathname; +{ + if (nfifo >= fifo_list_size - 1) + { + fifo_list_size += FIFO_INCR; + fifo_list = (struct temp_fifo *)xrealloc (fifo_list, + fifo_list_size * sizeof (struct temp_fifo)); + } + + fifo_list[nfifo].file = savestring (pathname); + nfifo++; +} + +void +unlink_fifo (i) + int i; +{ + if ((fifo_list[i].proc == -1) || (kill(fifo_list[i].proc, 0) == -1)) + { + unlink (fifo_list[i].file); + free (fifo_list[i].file); + fifo_list[i].file = (char *)NULL; + fifo_list[i].proc = -1; + } +} + +void +unlink_fifo_list () +{ + int saved, i, j; + + if (nfifo == 0) + return; + + for (i = saved = 0; i < nfifo; i++) + { + if ((fifo_list[i].proc == -1) || (kill(fifo_list[i].proc, 0) == -1)) + { + unlink (fifo_list[i].file); + free (fifo_list[i].file); + fifo_list[i].file = (char *)NULL; + fifo_list[i].proc = -1; + } + else + saved++; + } + + /* If we didn't remove some of the FIFOs, compact the list. */ + if (saved) + { + for (i = j = 0; i < nfifo; i++) + if (fifo_list[i].file) + { + fifo_list[j].file = fifo_list[i].file; + fifo_list[j].proc = fifo_list[i].proc; + j++; + } + nfifo = j; + } + else + nfifo = 0; +} + +/* Take LIST, which is a bitmap denoting active FIFOs in fifo_list + from some point in the past, and close all open FIFOs in fifo_list + that are not marked as active in LIST. If LIST is NULL, close + everything in fifo_list. LSIZE is the number of elements in LIST, in + case it's larger than fifo_list_size (size of fifo_list). */ +void +close_new_fifos (list, lsize) + char *list; + int lsize; +{ + int i; + + if (list == 0) + { + unlink_fifo_list (); + return; + } + + for (i = 0; i < lsize; i++) + if (list[i] == 0 && i < fifo_list_size && fifo_list[i].proc != -1) + unlink_fifo (i); + + for (i = lsize; i < fifo_list_size; i++) + unlink_fifo (i); +} + +int +fifos_pending () +{ + return nfifo; +} + +int +num_fifos () +{ + return nfifo; +} + +static char * +make_named_pipe () +{ + char *tname; + + tname = sh_mktmpname ("sh-np", MT_USERANDOM|MT_USETMPDIR); + if (mkfifo (tname, 0600) < 0) + { + free (tname); + return ((char *)NULL); + } + + add_fifo_list (tname); + return (tname); +} + +#else /* HAVE_DEV_FD */ + +/* DEV_FD_LIST is a bitmap of file descriptors attached to pipes the shell + has open to children. NFDS is a count of the number of bits currently + set in DEV_FD_LIST. TOTFDS is a count of the highest possible number + of open files. */ +static char *dev_fd_list = (char *)NULL; +static int nfds; +static int totfds; /* The highest possible number of open files. */ + +void +clear_fifo (i) + int i; +{ + if (dev_fd_list[i]) + { + dev_fd_list[i] = 0; + nfds--; + } +} + +void +clear_fifo_list () +{ + register int i; + + if (nfds == 0) + return; + + for (i = 0; nfds && i < totfds; i++) + clear_fifo (i); + + nfds = 0; +} + +char * +copy_fifo_list (sizep) + int *sizep; +{ + char *ret; + + if (nfds == 0 || totfds == 0) + { + if (sizep) + *sizep = 0; + return (char *)NULL; + } + + if (sizep) + *sizep = totfds; + ret = (char *)xmalloc (totfds); + return (memcpy (ret, dev_fd_list, totfds)); +} + +static void +add_fifo_list (fd) + int fd; +{ + if (dev_fd_list == 0 || fd >= totfds) + { + int ofds; + + ofds = totfds; + totfds = getdtablesize (); + if (totfds < 0 || totfds > 256) + totfds = 256; + if (fd >= totfds) + totfds = fd + 2; + + dev_fd_list = (char *)xrealloc (dev_fd_list, totfds); + memset (dev_fd_list + ofds, '\0', totfds - ofds); + } + + dev_fd_list[fd] = 1; + nfds++; +} + +int +fifos_pending () +{ + return 0; /* used for cleanup; not needed with /dev/fd */ +} + +int +num_fifos () +{ + return nfds; +} + +void +unlink_fifo (fd) + int fd; +{ + if (dev_fd_list[fd]) + { + close (fd); + dev_fd_list[fd] = 0; + nfds--; + } +} + +void +unlink_fifo_list () +{ + register int i; + + if (nfds == 0) + return; + + for (i = 0; nfds && i < totfds; i++) + unlink_fifo (i); + + nfds = 0; +} + +/* Take LIST, which is a snapshot copy of dev_fd_list from some point in + the past, and close all open fds in dev_fd_list that are not marked + as open in LIST. If LIST is NULL, close everything in dev_fd_list. + LSIZE is the number of elements in LIST, in case it's larger than + totfds (size of dev_fd_list). */ +void +close_new_fifos (list, lsize) + char *list; + int lsize; +{ + int i; + + if (list == 0) + { + unlink_fifo_list (); + return; + } + + for (i = 0; i < lsize; i++) + if (list[i] == 0 && i < totfds && dev_fd_list[i]) + unlink_fifo (i); + + for (i = lsize; i < totfds; i++) + unlink_fifo (i); +} + +#if defined (NOTDEF) +print_dev_fd_list () +{ + register int i; + + fprintf (stderr, "pid %ld: dev_fd_list:", (long)getpid ()); + fflush (stderr); + + for (i = 0; i < totfds; i++) + { + if (dev_fd_list[i]) + fprintf (stderr, " %d", i); + } + fprintf (stderr, "\n"); +} +#endif /* NOTDEF */ + +static char * +make_dev_fd_filename (fd) + int fd; +{ + char *ret, intbuf[INT_STRLEN_BOUND (int) + 1], *p; + + ret = (char *)xmalloc (sizeof (DEV_FD_PREFIX) + 8); + + strcpy (ret, DEV_FD_PREFIX); + p = inttostr (fd, intbuf, sizeof (intbuf)); + strcpy (ret + sizeof (DEV_FD_PREFIX) - 1, p); + + add_fifo_list (fd); + return (ret); +} + +#endif /* HAVE_DEV_FD */ + +/* Return a filename that will open a connection to the process defined by + executing STRING. HAVE_DEV_FD, if defined, means open a pipe and return + a filename in /dev/fd corresponding to a descriptor that is one of the + ends of the pipe. If not defined, we use named pipes on systems that have + them. Systems without /dev/fd and named pipes are out of luck. + + OPEN_FOR_READ_IN_CHILD, if 1, means open the named pipe for reading or + use the read end of the pipe and dup that file descriptor to fd 0 in + the child. If OPEN_FOR_READ_IN_CHILD is 0, we open the named pipe for + writing or use the write end of the pipe in the child, and dup that + file descriptor to fd 1 in the child. The parent does the opposite. */ + +static char * +process_substitute (string, open_for_read_in_child) + char *string; + int open_for_read_in_child; +{ + char *pathname; + int fd, result; + pid_t old_pid, pid; +#if defined (HAVE_DEV_FD) + int parent_pipe_fd, child_pipe_fd; + int fildes[2]; +#endif /* HAVE_DEV_FD */ +#if defined (JOB_CONTROL) + pid_t old_pipeline_pgrp; +#endif + + if (!string || !*string || wordexp_only) + return ((char *)NULL); + +#if !defined (HAVE_DEV_FD) + pathname = make_named_pipe (); +#else /* HAVE_DEV_FD */ + if (pipe (fildes) < 0) + { + sys_error (_("cannot make pipe for process substitution")); + return ((char *)NULL); + } + /* If OPEN_FOR_READ_IN_CHILD == 1, we want to use the write end of + the pipe in the parent, otherwise the read end. */ + parent_pipe_fd = fildes[open_for_read_in_child]; + child_pipe_fd = fildes[1 - open_for_read_in_child]; + /* Move the parent end of the pipe to some high file descriptor, to + avoid clashes with FDs used by the script. */ + parent_pipe_fd = move_to_high_fd (parent_pipe_fd, 1, 64); + + pathname = make_dev_fd_filename (parent_pipe_fd); +#endif /* HAVE_DEV_FD */ + + if (pathname == 0) + { + sys_error (_("cannot make pipe for process substitution")); + return ((char *)NULL); + } + + old_pid = last_made_pid; + +#if defined (JOB_CONTROL) + old_pipeline_pgrp = pipeline_pgrp; + pipeline_pgrp = shell_pgrp; + save_pipeline (1); +#endif /* JOB_CONTROL */ + + pid = make_child ((char *)NULL, 1); + if (pid == 0) + { + reset_terminating_signals (); /* XXX */ + free_pushed_string_input (); + /* Cancel traps, in trap.c. */ + restore_original_signals (); /* XXX - what about special builtins? bash-4.2 */ + setup_async_signals (); + subshell_environment |= SUBSHELL_COMSUB|SUBSHELL_PROCSUB; + } + +#if defined (JOB_CONTROL) + set_sigchld_handler (); + stop_making_children (); + /* XXX - should we only do this in the parent? (as in command subst) */ + pipeline_pgrp = old_pipeline_pgrp; +#endif /* JOB_CONTROL */ + + if (pid < 0) + { + sys_error (_("cannot make child for process substitution")); + free (pathname); +#if defined (HAVE_DEV_FD) + close (parent_pipe_fd); + close (child_pipe_fd); +#endif /* HAVE_DEV_FD */ + return ((char *)NULL); + } + + if (pid > 0) + { +#if defined (JOB_CONTROL) + restore_pipeline (1); +#endif + +#if !defined (HAVE_DEV_FD) + fifo_list[nfifo-1].proc = pid; +#endif + + last_made_pid = old_pid; + +#if defined (JOB_CONTROL) && defined (PGRP_PIPE) + close_pgrp_pipe (); +#endif /* JOB_CONTROL && PGRP_PIPE */ + +#if defined (HAVE_DEV_FD) + close (child_pipe_fd); +#endif /* HAVE_DEV_FD */ + + return (pathname); + } + + set_sigint_handler (); + +#if defined (JOB_CONTROL) + set_job_control (0); +#endif /* JOB_CONTROL */ + +#if !defined (HAVE_DEV_FD) + /* Open the named pipe in the child. */ + fd = open (pathname, open_for_read_in_child ? O_RDONLY : O_WRONLY); + if (fd < 0) + { + /* Two separate strings for ease of translation. */ + if (open_for_read_in_child) + sys_error (_("cannot open named pipe %s for reading"), pathname); + else + sys_error (_("cannot open named pipe %s for writing"), pathname); + + exit (127); + } + if (open_for_read_in_child) + { + if (sh_unset_nodelay_mode (fd) < 0) + { + sys_error (_("cannot reset nodelay mode for fd %d"), fd); + exit (127); + } + } +#else /* HAVE_DEV_FD */ + fd = child_pipe_fd; +#endif /* HAVE_DEV_FD */ + + if (dup2 (fd, open_for_read_in_child ? 0 : 1) < 0) + { + sys_error (_("cannot duplicate named pipe %s as fd %d"), pathname, + open_for_read_in_child ? 0 : 1); + exit (127); + } + + if (fd != (open_for_read_in_child ? 0 : 1)) + close (fd); + + /* Need to close any files that this process has open to pipes inherited + from its parent. */ + if (current_fds_to_close) + { + close_fd_bitmap (current_fds_to_close); + current_fds_to_close = (struct fd_bitmap *)NULL; + } + +#if defined (HAVE_DEV_FD) + /* Make sure we close the parent's end of the pipe and clear the slot + in the fd list so it is not closed later, if reallocated by, for + instance, pipe(2). */ + close (parent_pipe_fd); + dev_fd_list[parent_pipe_fd] = 0; +#endif /* HAVE_DEV_FD */ + + /* subshells shouldn't have this flag, which controls using the temporary + environment for variable lookups. */ + expanding_redir = 0; + + result = parse_and_execute (string, "process substitution", (SEVAL_NONINT|SEVAL_NOHIST)); + +#if !defined (HAVE_DEV_FD) + /* Make sure we close the named pipe in the child before we exit. */ + close (open_for_read_in_child ? 0 : 1); +#endif /* !HAVE_DEV_FD */ + + last_command_exit_value = result; + result = run_exit_trap (); + exit (result); + /*NOTREACHED*/ +} +#endif /* PROCESS_SUBSTITUTION */ + +/***********************************/ +/* */ +/* Command Substitution */ +/* */ +/***********************************/ + +static char * +read_comsub (fd, quoted, rflag) + int fd, quoted; + int *rflag; +{ + char *istring, buf[128], *bufp, *s; + int istring_index, istring_size, c, tflag, skip_ctlesc, skip_ctlnul; + ssize_t bufn; + + istring = (char *)NULL; + istring_index = istring_size = bufn = tflag = 0; + + for (skip_ctlesc = skip_ctlnul = 0, s = ifs_value; s && *s; s++) + skip_ctlesc |= *s == CTLESC, skip_ctlnul |= *s == CTLNUL; + + /* Read the output of the command through the pipe. This may need to be + changed to understand multibyte characters in the future. */ + while (1) + { + if (fd < 0) + break; + if (--bufn <= 0) + { + bufn = zread (fd, buf, sizeof (buf)); + if (bufn <= 0) + break; + bufp = buf; + } + c = *bufp++; + + if (c == 0) + { +#if 0 + internal_warning ("read_comsub: ignored null byte in input"); +#endif + continue; + } + + /* Add the character to ISTRING, possibly after resizing it. */ + RESIZE_MALLOCED_BUFFER (istring, istring_index, 2, istring_size, DEFAULT_ARRAY_SIZE); + + /* This is essentially quote_string inline */ + if ((quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) /* || c == CTLESC || c == CTLNUL */) + istring[istring_index++] = CTLESC; + /* Escape CTLESC and CTLNUL in the output to protect those characters + from the rest of the word expansions (word splitting and globbing.) + This is essentially quote_escapes inline. */ + else if (skip_ctlesc == 0 && c == CTLESC) + { + tflag |= W_HASCTLESC; + istring[istring_index++] = CTLESC; + } + else if ((skip_ctlnul == 0 && c == CTLNUL) || (c == ' ' && (ifs_value && *ifs_value == 0))) + istring[istring_index++] = CTLESC; + + istring[istring_index++] = c; + +#if 0 +#if defined (__CYGWIN__) + if (c == '\n' && istring_index > 1 && istring[istring_index - 2] == '\r') + { + istring_index--; + istring[istring_index - 1] = '\n'; + } +#endif +#endif + } + + if (istring) + istring[istring_index] = '\0'; + + /* If we read no output, just return now and save ourselves some + trouble. */ + if (istring_index == 0) + { + FREE (istring); + if (rflag) + *rflag = tflag; + return (char *)NULL; + } + + /* Strip trailing newlines from the output of the command. */ + if (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) + { + while (istring_index > 0) + { + if (istring[istring_index - 1] == '\n') + { + --istring_index; + + /* If the newline was quoted, remove the quoting char. */ + if (istring[istring_index - 1] == CTLESC) + --istring_index; + } + else + break; + } + istring[istring_index] = '\0'; + } + else + strip_trailing (istring, istring_index - 1, 1); + + if (rflag) + *rflag = tflag; + return istring; +} + +/* Perform command substitution on STRING. This returns a WORD_DESC * with the + contained string possibly quoted. */ +WORD_DESC * +command_substitute (string, quoted) + char *string; + int quoted; +{ + pid_t pid, old_pid, old_pipeline_pgrp, old_async_pid; + char *istring; + int result, fildes[2], function_value, pflags, rc, tflag; + WORD_DESC *ret; + + istring = (char *)NULL; + + /* Don't fork () if there is no need to. In the case of no command to + run, just return NULL. */ + if (!string || !*string || (string[0] == '\n' && !string[1])) + return ((WORD_DESC *)NULL); + + if (wordexp_only && read_but_dont_execute) + { + last_command_exit_value = EX_WEXPCOMSUB; + jump_to_top_level (EXITPROG); + } + + /* We're making the assumption here that the command substitution will + eventually run a command from the file system. Since we'll run + maybe_make_export_env in this subshell before executing that command, + the parent shell and any other shells it starts will have to remake + the environment. If we make it before we fork, other shells won't + have to. Don't bother if we have any temporary variable assignments, + though, because the export environment will be remade after this + command completes anyway, but do it if all the words to be expanded + are variable assignments. */ + if (subst_assign_varlist == 0 || garglist == 0) + maybe_make_export_env (); /* XXX */ + + /* Flags to pass to parse_and_execute() */ + pflags = (interactive && sourcelevel == 0) ? SEVAL_RESETLINE : 0; + + /* Pipe the output of executing STRING into the current shell. */ + if (pipe (fildes) < 0) + { + sys_error (_("cannot make pipe for command substitution")); + goto error_exit; + } + + old_pid = last_made_pid; +#if defined (JOB_CONTROL) + old_pipeline_pgrp = pipeline_pgrp; + /* Don't reset the pipeline pgrp if we're already a subshell in a pipeline. */ + if ((subshell_environment & SUBSHELL_PIPE) == 0) + pipeline_pgrp = shell_pgrp; + cleanup_the_pipeline (); +#endif /* JOB_CONTROL */ + + old_async_pid = last_asynchronous_pid; + pid = make_child ((char *)NULL, subshell_environment&SUBSHELL_ASYNC); + last_asynchronous_pid = old_async_pid; + + if (pid == 0) + { + /* Reset the signal handlers in the child, but don't free the + trap strings. Set a flag noting that we have to free the + trap strings if we run trap to change a signal disposition. */ + reset_signal_handlers (); + subshell_environment |= SUBSHELL_RESETTRAP; + } + +#if defined (JOB_CONTROL) + /* XXX DO THIS ONLY IN PARENT ? XXX */ + set_sigchld_handler (); + stop_making_children (); + if (pid != 0) + pipeline_pgrp = old_pipeline_pgrp; +#else + stop_making_children (); +#endif /* JOB_CONTROL */ + + if (pid < 0) + { + sys_error (_("cannot make child for command substitution")); + error_exit: + + last_made_pid = old_pid; + + FREE (istring); + close (fildes[0]); + close (fildes[1]); + return ((WORD_DESC *)NULL); + } + + if (pid == 0) + { + set_sigint_handler (); /* XXX */ + + free_pushed_string_input (); + + if (dup2 (fildes[1], 1) < 0) + { + sys_error (_("command_substitute: cannot duplicate pipe as fd 1")); + exit (EXECUTION_FAILURE); + } + + /* If standard output is closed in the parent shell + (such as after `exec >&-'), file descriptor 1 will be + the lowest available file descriptor, and end up in + fildes[0]. This can happen for stdin and stderr as well, + but stdout is more important -- it will cause no output + to be generated from this command. */ + if ((fildes[1] != fileno (stdin)) && + (fildes[1] != fileno (stdout)) && + (fildes[1] != fileno (stderr))) + close (fildes[1]); + + if ((fildes[0] != fileno (stdin)) && + (fildes[0] != fileno (stdout)) && + (fildes[0] != fileno (stderr))) + close (fildes[0]); + +#ifdef __CYGWIN__ + /* Let stdio know the fd may have changed from text to binary mode, and + make sure to preserve stdout line buffering. */ + freopen (NULL, "w", stdout); + sh_setlinebuf (stdout); +#endif /* __CYGWIN__ */ + + /* The currently executing shell is not interactive. */ + interactive = 0; + + /* This is a subshell environment. */ + subshell_environment |= SUBSHELL_COMSUB; + + /* Many shells do not appear to inherit the -v option for command + substitutions. */ + change_flag ('v', FLAG_OFF); + + /* When not in POSIX mode, command substitution does not inherit + the -e flag. */ + if (posixly_correct == 0) + { + builtin_ignoring_errexit = 0; + change_flag ('e', FLAG_OFF); + set_shellopts (); + } + + remove_quoted_escapes (string); + + startup_state = 2; /* see if we can avoid a fork */ + /* Give command substitution a place to jump back to on failure, + so we don't go back up to main (). */ + result = setjmp_nosigs (top_level); + + /* If we're running a command substitution inside a shell function, + trap `return' so we don't return from the function in the subshell + and go off to never-never land. */ + if (result == 0 && return_catch_flag) + function_value = setjmp_nosigs (return_catch); + else + function_value = 0; + + if (result == ERREXIT) + rc = last_command_exit_value; + else if (result == EXITPROG) + rc = last_command_exit_value; + else if (result) + rc = EXECUTION_FAILURE; + else if (function_value) + rc = return_catch_value; + else + { + subshell_level++; + rc = parse_and_execute (string, "command substitution", pflags|SEVAL_NOHIST); + subshell_level--; + } + + last_command_exit_value = rc; + rc = run_exit_trap (); +#if defined (PROCESS_SUBSTITUTION) + unlink_fifo_list (); +#endif + exit (rc); + } + else + { +#if defined (JOB_CONTROL) && defined (PGRP_PIPE) + close_pgrp_pipe (); +#endif /* JOB_CONTROL && PGRP_PIPE */ + + close (fildes[1]); + + tflag = 0; + istring = read_comsub (fildes[0], quoted, &tflag); + + close (fildes[0]); + + current_command_subst_pid = pid; + last_command_exit_value = wait_for (pid); + last_command_subst_pid = pid; + last_made_pid = old_pid; + +#if defined (JOB_CONTROL) + /* If last_command_exit_value > 128, then the substituted command + was terminated by a signal. If that signal was SIGINT, then send + SIGINT to ourselves. This will break out of loops, for instance. */ + if (last_command_exit_value == (128 + SIGINT) && last_command_exit_signal == SIGINT) + kill (getpid (), SIGINT); + + /* wait_for gives the terminal back to shell_pgrp. If some other + process group should have it, give it away to that group here. + pipeline_pgrp is non-zero only while we are constructing a + pipeline, so what we are concerned about is whether or not that + pipeline was started in the background. A pipeline started in + the background should never get the tty back here. */ + if (interactive && pipeline_pgrp != (pid_t)0 && (subshell_environment & SUBSHELL_ASYNC) == 0) + give_terminal_to (pipeline_pgrp, 0); +#endif /* JOB_CONTROL */ + + ret = alloc_word_desc (); + ret->word = istring; + ret->flags = tflag; + + return ret; + } +} + +/******************************************************** + * * + * Utility functions for parameter expansion * + * * + ********************************************************/ + +#if defined (ARRAY_VARS) + +static arrayind_t +array_length_reference (s) + char *s; +{ + int len; + arrayind_t ind; + char *akey; + char *t, c; + ARRAY *array; + HASH_TABLE *h; + SHELL_VAR *var; + + var = array_variable_part (s, &t, &len); + + /* If unbound variables should generate an error, report one and return + failure. */ + if ((var == 0 || invisible_p (var) || (assoc_p (var) == 0 && array_p (var) == 0)) && unbound_vars_is_error) + { + c = *--t; + *t = '\0'; + last_command_exit_value = EXECUTION_FAILURE; + err_unboundvar (s); + *t = c; + return (-1); + } + else if (var == 0 || invisible_p (var)) + return 0; + + /* We support a couple of expansions for variables that are not arrays. + We'll return the length of the value for v[0], and 1 for v[@] or + v[*]. Return 0 for everything else. */ + + array = array_p (var) ? array_cell (var) : (ARRAY *)NULL; + h = assoc_p (var) ? assoc_cell (var) : (HASH_TABLE *)NULL; + + if (ALL_ELEMENT_SUB (t[0]) && t[1] == ']') + { + if (assoc_p (var)) + return (h ? assoc_num_elements (h) : 0); + else if (array_p (var)) + return (array ? array_num_elements (array) : 0); + else + return (var_isset (var) ? 1 : 0); + } + + if (assoc_p (var)) + { + t[len - 1] = '\0'; + akey = expand_assignment_string_to_string (t, 0); /* [ */ + t[len - 1] = ']'; + if (akey == 0 || *akey == 0) + { + err_badarraysub (t); + FREE (akey); + return (-1); + } + t = assoc_reference (assoc_cell (var), akey); + free (akey); + } + else + { + ind = array_expand_index (var, t, len); + /* negative subscripts to indexed arrays count back from end */ + if (var && array_p (var) && ind < 0) + ind = array_max_index (array_cell (var)) + 1 + ind; + if (ind < 0) + { + err_badarraysub (t); + return (-1); + } + if (array_p (var)) + t = array_reference (array, ind); + else + t = (ind == 0) ? value_cell (var) : (char *)NULL; + } + + len = MB_STRLEN (t); + return (len); +} +#endif /* ARRAY_VARS */ + +static int +valid_brace_expansion_word (name, var_is_special) + char *name; + int var_is_special; +{ + if (DIGIT (*name) && all_digits (name)) + return 1; + else if (var_is_special) + return 1; +#if defined (ARRAY_VARS) + else if (valid_array_reference (name)) + return 1; +#endif /* ARRAY_VARS */ + else if (legal_identifier (name)) + return 1; + else + return 0; +} + +static int +chk_atstar (name, quoted, quoted_dollar_atp, contains_dollar_at) + char *name; + int quoted; + int *quoted_dollar_atp, *contains_dollar_at; +{ + char *temp1; + + if (name == 0) + { + if (quoted_dollar_atp) + *quoted_dollar_atp = 0; + if (contains_dollar_at) + *contains_dollar_at = 0; + return 0; + } + + /* check for $@ and $* */ + if (name[0] == '@' && name[1] == 0) + { + if ((quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) && quoted_dollar_atp) + *quoted_dollar_atp = 1; + if (contains_dollar_at) + *contains_dollar_at = 1; + return 1; + } + else if (name[0] == '*' && name[1] == '\0' && quoted == 0) + { + if (contains_dollar_at) + *contains_dollar_at = 1; + return 1; + } + + /* Now check for ${array[@]} and ${array[*]} */ +#if defined (ARRAY_VARS) + else if (valid_array_reference (name)) + { + temp1 = mbschr (name, '['); + if (temp1 && temp1[1] == '@' && temp1[2] == ']') + { + if ((quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) && quoted_dollar_atp) + *quoted_dollar_atp = 1; + if (contains_dollar_at) + *contains_dollar_at = 1; + return 1; + } /* [ */ + /* ${array[*]}, when unquoted, should be treated like ${array[@]}, + which should result in separate words even when IFS is unset. */ + if (temp1 && temp1[1] == '*' && temp1[2] == ']' && quoted == 0) + { + if (contains_dollar_at) + *contains_dollar_at = 1; + return 1; + } + } +#endif + return 0; +} + +/* Parameter expand NAME, and return a new string which is the expansion, + or NULL if there was no expansion. + VAR_IS_SPECIAL is non-zero if NAME is one of the special variables in + the shell, e.g., "@", "$", "*", etc. QUOTED, if non-zero, means that + NAME was found inside of a double-quoted expression. */ +static WORD_DESC * +parameter_brace_expand_word (name, var_is_special, quoted, pflags, indp) + char *name; + int var_is_special, quoted, pflags; + arrayind_t *indp; +{ + WORD_DESC *ret; + char *temp, *tt; + intmax_t arg_index; + SHELL_VAR *var; + int atype, rflags; + arrayind_t ind; + + ret = 0; + temp = 0; + rflags = 0; + + if (indp) + *indp = INTMAX_MIN; + + /* Handle multiple digit arguments, as in ${11}. */ + if (legal_number (name, &arg_index)) + { + tt = get_dollar_var_value (arg_index); + if (tt) + temp = (*tt && (quoted & (Q_DOUBLE_QUOTES|Q_HERE_DOCUMENT))) + ? quote_string (tt) + : quote_escapes (tt); + else + temp = (char *)NULL; + FREE (tt); + } + else if (var_is_special) /* ${@} */ + { + int sindex; + tt = (char *)xmalloc (2 + strlen (name)); + tt[sindex = 0] = '$'; + strcpy (tt + 1, name); + + ret = param_expand (tt, &sindex, quoted, (int *)NULL, (int *)NULL, + (int *)NULL, (int *)NULL, pflags); + free (tt); + } +#if defined (ARRAY_VARS) + else if (valid_array_reference (name)) + { +expand_arrayref: + /* XXX - does this leak if name[@] or name[*]? */ + if (pflags & PF_ASSIGNRHS) + { + var = array_variable_part (name, &tt, (int *)0); + if (ALL_ELEMENT_SUB (tt[0]) && tt[1] == ']') + { + /* Only treat as double quoted if array variable */ + if (var && (array_p (var) || assoc_p (var))) + temp = array_value (name, quoted|Q_DOUBLE_QUOTES, 0, &atype, &ind); + else + temp = array_value (name, quoted, 0, &atype, &ind); + } + else + temp = array_value (name, quoted, 0, &atype, &ind); + } + else + temp = array_value (name, quoted, 0, &atype, &ind); + if (atype == 0 && temp) + { + temp = (*temp && (quoted & (Q_DOUBLE_QUOTES|Q_HERE_DOCUMENT))) + ? quote_string (temp) + : quote_escapes (temp); + rflags |= W_ARRAYIND; + if (indp) + *indp = ind; + } + else if (atype == 1 && temp && QUOTED_NULL (temp) && (quoted & (Q_DOUBLE_QUOTES|Q_HERE_DOCUMENT))) + rflags |= W_HASQUOTEDNULL; + } +#endif + else if (var = find_variable (name)) + { + if (var_isset (var) && invisible_p (var) == 0) + { +#if defined (ARRAY_VARS) + if (assoc_p (var)) + temp = assoc_reference (assoc_cell (var), "0"); + else if (array_p (var)) + temp = array_reference (array_cell (var), 0); + else + temp = value_cell (var); +#else + temp = value_cell (var); +#endif + + if (temp) + temp = (*temp && (quoted & (Q_DOUBLE_QUOTES|Q_HERE_DOCUMENT))) + ? quote_string (temp) + : quote_escapes (temp); + } + else + temp = (char *)NULL; + } + else if (var = find_variable_last_nameref (name)) + { + temp = nameref_cell (var); +#if defined (ARRAY_VARS) + /* Handle expanding nameref whose value is x[n] */ + if (temp && *temp && valid_array_reference (temp)) + { + name = temp; + goto expand_arrayref; + } + else +#endif + /* y=2 ; typeset -n x=y; echo ${x} is not the same as echo ${2} in ksh */ + if (temp && *temp && legal_identifier (temp) == 0) + { + last_command_exit_value = EXECUTION_FAILURE; + report_error (_("%s: invalid variable name for name reference"), temp); + temp = &expand_param_error; + } + else + temp = (char *)NULL; + } + else + temp = (char *)NULL; + + if (ret == 0) + { + ret = alloc_word_desc (); + ret->word = temp; + ret->flags |= rflags; + } + return ret; +} + +static char * +parameter_brace_find_indir (name, var_is_special, quoted, find_nameref) + char *name; + int var_is_special, quoted, find_nameref; +{ + char *temp, *t; + WORD_DESC *w; + SHELL_VAR *v; + + if (find_nameref && var_is_special == 0 && (v = find_variable_last_nameref (name)) && + nameref_p (v) && (t = nameref_cell (v)) && *t) + return (savestring (t)); + + /* If var_is_special == 0, and name is not an array reference, this does + more expansion than necessary. It should really look up the variable's + value and not try to expand it. */ + w = parameter_brace_expand_word (name, var_is_special, quoted, PF_IGNUNBOUND, 0); + t = w->word; + /* Have to dequote here if necessary */ + if (t) + { + temp = (quoted & (Q_DOUBLE_QUOTES|Q_HERE_DOCUMENT)) + ? dequote_string (t) + : dequote_escapes (t); + free (t); + t = temp; + } + dispose_word_desc (w); + + return t; +} + +/* Expand an indirect reference to a variable: ${!NAME} expands to the + value of the variable whose name is the value of NAME. */ +static WORD_DESC * +parameter_brace_expand_indir (name, var_is_special, quoted, quoted_dollar_atp, contains_dollar_at) + char *name; + int var_is_special, quoted; + int *quoted_dollar_atp, *contains_dollar_at; +{ + char *temp, *t; + WORD_DESC *w; + SHELL_VAR *v; + + /* See if it's a nameref first, behave in ksh93-compatible fashion. + There is at least one incompatibility: given ${!foo[0]} where foo=bar, + bash performs an indirect lookup on foo[0] and expands the result; + ksh93 expands bar[0]. We could do that here -- there are enough usable + primitives to do that -- but do not at this point. */ + if (var_is_special == 0 && (v = find_variable_last_nameref (name))) + { + if (nameref_p (v) && (t = nameref_cell (v)) && *t) + { + w = alloc_word_desc (); + w->word = savestring (t); + w->flags = 0; + return w; + } + } + + t = parameter_brace_find_indir (name, var_is_special, quoted, 0); + + chk_atstar (t, quoted, quoted_dollar_atp, contains_dollar_at); + if (t == 0) + return (WORD_DESC *)NULL; + + w = parameter_brace_expand_word (t, SPECIAL_VAR(t, 0), quoted, 0, 0); + free (t); + + return w; +} + +/* Expand the right side of a parameter expansion of the form ${NAMEcVALUE}, + depending on the value of C, the separating character. C can be one of + "-", "+", or "=". QUOTED is true if the entire brace expression occurs + between double quotes. */ +static WORD_DESC * +parameter_brace_expand_rhs (name, value, c, quoted, qdollaratp, hasdollarat) + char *name, *value; + int c, quoted, *qdollaratp, *hasdollarat; +{ + WORD_DESC *w; + WORD_LIST *l; + char *t, *t1, *temp, *vname; + int hasdol; + + /* If the entire expression is between double quotes, we want to treat + the value as a double-quoted string, with the exception that we strip + embedded unescaped double quotes (for sh backwards compatibility). */ + if ((quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) && *value) + { + hasdol = 0; + temp = string_extract_double_quoted (value, &hasdol, 1); + } + else + temp = value; + + w = alloc_word_desc (); + hasdol = 0; + /* XXX was 0 not quoted */ + l = *temp ? expand_string_for_rhs (temp, quoted, &hasdol, (int *)NULL) + : (WORD_LIST *)0; + if (hasdollarat) + *hasdollarat = hasdol || (l && l->next); + if (temp != value) + free (temp); + if (l) + { + /* The expansion of TEMP returned something. We need to treat things + slightly differently if HASDOL is non-zero. If we have "$@", the + individual words have already been quoted. We need to turn them + into a string with the words separated by the first character of + $IFS without any additional quoting, so string_list_dollar_at won't + do the right thing. We use string_list_dollar_star instead. */ + temp = (hasdol || l->next) ? string_list_dollar_star (l) : string_list (l); + + /* If l->next is not null, we know that TEMP contained "$@", since that + is the only expansion that creates more than one word. */ + if (qdollaratp && ((hasdol && quoted) || l->next)) + *qdollaratp = 1; + /* If we have a quoted null result (QUOTED_NULL(temp)) and the word is + a quoted null (l->next == 0 && QUOTED_NULL(l->word->word)), the + flags indicate it (l->word->flags & W_HASQUOTEDNULL), and the + expansion is quoted (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) + (which is more paranoia than anything else), we need to return the + quoted null string and set the flags to indicate it. */ + if (l->next == 0 && (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) && QUOTED_NULL (temp) && QUOTED_NULL (l->word->word) && (l->word->flags & W_HASQUOTEDNULL)) + { + w->flags |= W_HASQUOTEDNULL; + } + dispose_words (l); + } + else if ((quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) && hasdol) + { +itrace("parameter_brace_expand_rhs: expanded to nothing with hasdol"); + /* Posix interp 221 changed the rules on this. The idea is that + something like "$xxx$@" should expand the same as "${foo-$xxx$@}" + when foo and xxx are unset. The problem is that it's not in any + way backwards compatible and few other shells do it. We're going + to try and split the difference (heh) a little bit here. */ + /* hasdol == 1 means we saw a quoted dollar at. */ + + /* The brace expansion occurred between double quotes and there was + a $@ in TEMP. It does not matter if the $@ is quoted, as long as + it does not expand to anything. In this case, we want to return + a quoted empty string. */ + temp = make_quoted_char ('\0'); + w->flags |= W_HASQUOTEDNULL; + } + else + temp = (char *)NULL; + + if (c == '-' || c == '+') + { + w->word = temp; + return w; + } + + /* c == '=' */ + t = temp ? savestring (temp) : savestring (""); + t1 = dequote_string (t); + free (t); + + /* bash-4.4/5.0 */ + vname = name; + if (*name == '!' && + (legal_variable_starter ((unsigned char)name[1]) || DIGIT (name[1]) || VALID_INDIR_PARAM (name[1]))) + { + vname = parameter_brace_find_indir (name + 1, SPECIAL_VAR (name, 1), quoted, 1); + if (vname == 0 || *vname == 0) + { + report_error (_("%s: invalid indirect expansion"), name); + free (vname); + dispose_word (w); + return &expand_wdesc_error; + } + if (legal_identifier (vname) == 0) + { + report_error (_("%s: invalid variable name"), vname); + free (vname); + dispose_word (w); + return &expand_wdesc_error; + } + } + +#if defined (ARRAY_VARS) + if (valid_array_reference (vname)) + assign_array_element (vname, t1, 0); + else +#endif /* ARRAY_VARS */ + bind_variable (vname, t1, 0); + + stupidly_hack_special_variables (vname); + + if (vname != name) + free (vname); + + /* From Posix group discussion Feb-March 2010. Issue 7 0000221 */ + free (temp); + + w->word = t1; + return w; +} + +/* Deal with the right hand side of a ${name:?value} expansion in the case + that NAME is null or not set. If VALUE is non-null it is expanded and + used as the error message to print, otherwise a standard message is + printed. */ +static void +parameter_brace_expand_error (name, value) + char *name, *value; +{ + WORD_LIST *l; + char *temp; + + last_command_exit_value = EXECUTION_FAILURE; /* ensure it's non-zero */ + if (value && *value) + { + l = expand_string (value, 0); + temp = string_list (l); + report_error ("%s: %s", name, temp ? temp : ""); /* XXX was value not "" */ + FREE (temp); + dispose_words (l); + } + else + report_error (_("%s: parameter null or not set"), name); + + /* Free the data we have allocated during this expansion, since we + are about to longjmp out. */ + free (name); + FREE (value); +} + +/* Return 1 if NAME is something for which parameter_brace_expand_length is + OK to do. */ +static int +valid_length_expression (name) + char *name; +{ + return (name[1] == '\0' || /* ${#} */ + ((sh_syntaxtab[(unsigned char) name[1]] & CSPECVAR) && name[2] == '\0') || /* special param */ + (DIGIT (name[1]) && all_digits (name + 1)) || /* ${#11} */ +#if defined (ARRAY_VARS) + valid_array_reference (name + 1) || /* ${#a[7]} */ +#endif + legal_identifier (name + 1)); /* ${#PS1} */ +} + +/* Handle the parameter brace expansion that requires us to return the + length of a parameter. */ +static intmax_t +parameter_brace_expand_length (name) + char *name; +{ + char *t, *newname; + intmax_t number, arg_index; + WORD_LIST *list; +#if defined (ARRAY_VARS) + SHELL_VAR *var; +#endif + + if (name[1] == '\0') /* ${#} */ + number = number_of_args (); + else if ((name[1] == '@' || name[1] == '*') && name[2] == '\0') /* ${#@}, ${#*} */ + number = number_of_args (); + else if ((sh_syntaxtab[(unsigned char) name[1]] & CSPECVAR) && name[2] == '\0') + { + /* Take the lengths of some of the shell's special parameters. */ + switch (name[1]) + { + case '-': + t = which_set_flags (); + break; + case '?': + t = itos (last_command_exit_value); + break; + case '$': + t = itos (dollar_dollar_pid); + break; + case '!': + if (last_asynchronous_pid == NO_PID) + t = (char *)NULL; /* XXX - error if set -u set? */ + else + t = itos (last_asynchronous_pid); + break; + case '#': + t = itos (number_of_args ()); + break; + } + number = STRLEN (t); + FREE (t); + } +#if defined (ARRAY_VARS) + else if (valid_array_reference (name + 1)) + number = array_length_reference (name + 1); +#endif /* ARRAY_VARS */ + else + { + number = 0; + + if (legal_number (name + 1, &arg_index)) /* ${#1} */ + { + t = get_dollar_var_value (arg_index); + if (t == 0 && unbound_vars_is_error) + return INTMAX_MIN; + number = MB_STRLEN (t); + FREE (t); + } +#if defined (ARRAY_VARS) + else if ((var = find_variable (name + 1)) && (invisible_p (var) == 0) && (array_p (var) || assoc_p (var))) + { + if (assoc_p (var)) + t = assoc_reference (assoc_cell (var), "0"); + else + t = array_reference (array_cell (var), 0); + if (t == 0 && unbound_vars_is_error) + return INTMAX_MIN; + number = MB_STRLEN (t); + } +#endif + else /* ${#PS1} */ + { + newname = savestring (name); + newname[0] = '$'; + list = expand_string (newname, Q_DOUBLE_QUOTES); + t = list ? string_list (list) : (char *)NULL; + free (newname); + if (list) + dispose_words (list); + + number = t ? MB_STRLEN (t) : 0; + FREE (t); + } + } + + return (number); +} + +/* Skip characters in SUBSTR until DELIM. SUBSTR is an arithmetic expression, + so we do some ad-hoc parsing of an arithmetic expression to find + the first DELIM, instead of using strchr(3). Two rules: + 1. If the substring contains a `(', read until closing `)'. + 2. If the substring contains a `?', read past one `:' for each `?'. +*/ + +static char * +skiparith (substr, delim) + char *substr; + int delim; +{ + size_t sublen; + int skipcol, pcount, i; + DECLARE_MBSTATE; + + sublen = strlen (substr); + i = skipcol = pcount = 0; + while (substr[i]) + { + /* Balance parens */ + if (substr[i] == LPAREN) + { + pcount++; + i++; + continue; + } + if (substr[i] == RPAREN && pcount) + { + pcount--; + i++; + continue; + } + if (pcount) + { + ADVANCE_CHAR (substr, sublen, i); + continue; + } + + /* Skip one `:' for each `?' */ + if (substr[i] == ':' && skipcol) + { + skipcol--; + i++; + continue; + } + if (substr[i] == delim) + break; + if (substr[i] == '?') + { + skipcol++; + i++; + continue; + } + ADVANCE_CHAR (substr, sublen, i); + } + + return (substr + i); +} + +/* Verify and limit the start and end of the desired substring. If + VTYPE == 0, a regular shell variable is being used; if it is 1, + then the positional parameters are being used; if it is 2, then + VALUE is really a pointer to an array variable that should be used. + Return value is 1 if both values were OK, 0 if there was a problem + with an invalid expression, or -1 if the values were out of range. */ +static int +verify_substring_values (v, value, substr, vtype, e1p, e2p) + SHELL_VAR *v; + char *value, *substr; + int vtype; + intmax_t *e1p, *e2p; +{ + char *t, *temp1, *temp2; + arrayind_t len; + int expok; +#if defined (ARRAY_VARS) + ARRAY *a; + HASH_TABLE *h; +#endif + + /* duplicate behavior of strchr(3) */ + t = skiparith (substr, ':'); + if (*t && *t == ':') + *t = '\0'; + else + t = (char *)0; + + temp1 = expand_arith_string (substr, Q_DOUBLE_QUOTES); + *e1p = evalexp (temp1, &expok); + free (temp1); + if (expok == 0) + return (0); + + len = -1; /* paranoia */ + switch (vtype) + { + case VT_VARIABLE: + case VT_ARRAYMEMBER: + len = MB_STRLEN (value); + break; + case VT_POSPARMS: + len = number_of_args () + 1; + if (*e1p == 0) + len++; /* add one arg if counting from $0 */ + break; +#if defined (ARRAY_VARS) + case VT_ARRAYVAR: + /* For arrays, the first value deals with array indices. Negative + offsets count from one past the array's maximum index. Associative + arrays treat the number of elements as the maximum index. */ + if (assoc_p (v)) + { + h = assoc_cell (v); + len = assoc_num_elements (h) + (*e1p < 0); + } + else + { + a = (ARRAY *)value; + len = array_max_index (a) + (*e1p < 0); /* arrays index from 0 to n - 1 */ + } + break; +#endif + } + + if (len == -1) /* paranoia */ + return -1; + + if (*e1p < 0) /* negative offsets count from end */ + *e1p += len; + + if (*e1p > len || *e1p < 0) + return (-1); + +#if defined (ARRAY_VARS) + /* For arrays, the second offset deals with the number of elements. */ + if (vtype == VT_ARRAYVAR) + len = assoc_p (v) ? assoc_num_elements (h) : array_num_elements (a); +#endif + + if (t) + { + t++; + temp2 = savestring (t); + temp1 = expand_arith_string (temp2, Q_DOUBLE_QUOTES); + free (temp2); + t[-1] = ':'; + *e2p = evalexp (temp1, &expok); + free (temp1); + if (expok == 0) + return (0); +#if 1 + if ((vtype == VT_ARRAYVAR || vtype == VT_POSPARMS) && *e2p < 0) +#else + /* bash-4.3: allow positional parameter length < 0 to count backwards + from end of positional parameters */ + if (vtype == VT_ARRAYVAR && *e2p < 0) +#endif + { + internal_error (_("%s: substring expression < 0"), t); + return (0); + } +#if defined (ARRAY_VARS) + /* In order to deal with sparse arrays, push the intelligence about how + to deal with the number of elements desired down to the array- + specific functions. */ + if (vtype != VT_ARRAYVAR) +#endif + { + if (*e2p < 0) + { + *e2p += len; + if (*e2p < 0 || *e2p < *e1p) + { + internal_error (_("%s: substring expression < 0"), t); + return (0); + } + } + else + *e2p += *e1p; /* want E2 chars starting at E1 */ + if (*e2p > len) + *e2p = len; + } + } + else + *e2p = len; + + return (1); +} + +/* Return the type of variable specified by VARNAME (simple variable, + positional param, or array variable). Also return the value specified + by VARNAME (value of a variable or a reference to an array element). + QUOTED is the standard description of quoting state, using Q_* defines. + FLAGS is currently a set of flags to pass to array_value. If IND is + non-null and not INTMAX_MIN, and FLAGS includes AV_USEIND, IND is + passed to array_value so the array index is not computed again. + If this returns VT_VARIABLE, the caller assumes that CTLESC and CTLNUL + characters in the value are quoted with CTLESC and takes appropriate + steps. For convenience, *VALP is set to the dequoted VALUE. */ +static int +get_var_and_type (varname, value, ind, quoted, flags, varp, valp) + char *varname, *value; + arrayind_t ind; + int quoted, flags; + SHELL_VAR **varp; + char **valp; +{ + int vtype, want_indir; + char *temp, *vname; + WORD_DESC *wd; +#if defined (ARRAY_VARS) + SHELL_VAR *v; +#endif + arrayind_t lind; + + want_indir = *varname == '!' && + (legal_variable_starter ((unsigned char)varname[1]) || DIGIT (varname[1]) + || VALID_INDIR_PARAM (varname[1])); + if (want_indir) + vname = parameter_brace_find_indir (varname+1, SPECIAL_VAR (varname, 1), quoted, 1); + /* XXX - what if vname == 0 || *vname == 0 ? */ + else + vname = varname; + + /* This sets vtype to VT_VARIABLE or VT_POSPARMS */ + vtype = (vname[0] == '@' || vname[0] == '*') && vname[1] == '\0'; + if (vtype == VT_POSPARMS && vname[0] == '*') + vtype |= VT_STARSUB; + *varp = (SHELL_VAR *)NULL; + +#if defined (ARRAY_VARS) + if (valid_array_reference (vname)) + { + v = array_variable_part (vname, &temp, (int *)0); + /* If we want to signal array_value to use an already-computed index, + set LIND to that index */ + lind = (ind != INTMAX_MIN && (flags & AV_USEIND)) ? ind : 0; + if (v && invisible_p (v)) + { + vtype = VT_ARRAYMEMBER; + *varp = (SHELL_VAR *)NULL; + *valp = (char *)NULL; + } + if (v && (array_p (v) || assoc_p (v))) + { /* [ */ + if (ALL_ELEMENT_SUB (temp[0]) && temp[1] == ']') + { + /* Callers have to differentiate between indexed and associative */ + vtype = VT_ARRAYVAR; + if (temp[0] == '*') + vtype |= VT_STARSUB; + *valp = array_p (v) ? (char *)array_cell (v) : (char *)assoc_cell (v); + } + else + { + vtype = VT_ARRAYMEMBER; + *valp = array_value (vname, Q_DOUBLE_QUOTES, flags, (int *)NULL, &lind); + } + *varp = v; + } + else if (v && (ALL_ELEMENT_SUB (temp[0]) && temp[1] == ']')) + { + vtype = VT_VARIABLE; + *varp = v; + if (quoted & (Q_DOUBLE_QUOTES|Q_HERE_DOCUMENT)) + *valp = dequote_string (value); + else + *valp = dequote_escapes (value); + } + else + { + vtype = VT_ARRAYMEMBER; + *varp = v; + *valp = array_value (vname, Q_DOUBLE_QUOTES, flags, (int *)NULL, &lind); + } + } + else if ((v = find_variable (vname)) && (invisible_p (v) == 0) && (assoc_p (v) || array_p (v))) + { + vtype = VT_ARRAYMEMBER; + *varp = v; + *valp = assoc_p (v) ? assoc_reference (assoc_cell (v), "0") : array_reference (array_cell (v), 0); + } + else +#endif + { + if (value && vtype == VT_VARIABLE) + { + if (quoted & (Q_DOUBLE_QUOTES|Q_HERE_DOCUMENT)) + *valp = dequote_string (value); + else + *valp = dequote_escapes (value); + } + else + *valp = value; + } + + if (want_indir) + free (vname); + + return vtype; +} + +/******************************************************/ +/* */ +/* Functions to extract substrings of variable values */ +/* */ +/******************************************************/ + +#if defined (HANDLE_MULTIBYTE) +/* Character-oriented rather than strictly byte-oriented substrings. S and + E, rather being strict indices into STRING, indicate character (possibly + multibyte character) positions that require calculation. + Used by the ${param:offset[:length]} expansion. */ +static char * +mb_substring (string, s, e) + char *string; + int s, e; +{ + char *tt; + int start, stop, i, slen; + DECLARE_MBSTATE; + + start = 0; + /* Don't need string length in ADVANCE_CHAR unless multibyte chars possible. */ + slen = (MB_CUR_MAX > 1) ? STRLEN (string) : 0; + + i = s; + while (string[start] && i--) + ADVANCE_CHAR (string, slen, start); + stop = start; + i = e - s; + while (string[stop] && i--) + ADVANCE_CHAR (string, slen, stop); + tt = substring (string, start, stop); + return tt; +} +#endif + +/* Process a variable substring expansion: ${name:e1[:e2]}. If VARNAME + is `@', use the positional parameters; otherwise, use the value of + VARNAME. If VARNAME is an array variable, use the array elements. */ + +static char * +parameter_brace_substring (varname, value, ind, substr, quoted, flags) + char *varname, *value; + int ind; + char *substr; + int quoted, flags; +{ + intmax_t e1, e2; + int vtype, r, starsub; + char *temp, *val, *tt, *oname; + SHELL_VAR *v; + + if (value == 0 && ((varname[0] != '@' && varname[0] != '*') || varname[1])) + return ((char *)NULL); + + oname = this_command_name; + this_command_name = varname; + + vtype = get_var_and_type (varname, value, ind, quoted, flags, &v, &val); + if (vtype == -1) + { + this_command_name = oname; + return ((char *)NULL); + } + + starsub = vtype & VT_STARSUB; + vtype &= ~VT_STARSUB; + + r = verify_substring_values (v, val, substr, vtype, &e1, &e2); + this_command_name = oname; + if (r <= 0) + { + if (vtype == VT_VARIABLE) + FREE (val); + return ((r == 0) ? &expand_param_error : (char *)NULL); + } + + switch (vtype) + { + case VT_VARIABLE: + case VT_ARRAYMEMBER: +#if defined (HANDLE_MULTIBYTE) + if (MB_CUR_MAX > 1) + tt = mb_substring (val, e1, e2); + else +#endif + tt = substring (val, e1, e2); + + if (vtype == VT_VARIABLE) + FREE (val); + if (quoted & (Q_DOUBLE_QUOTES|Q_HERE_DOCUMENT)) + temp = quote_string (tt); + else + temp = tt ? quote_escapes (tt) : (char *)NULL; + FREE (tt); + break; + case VT_POSPARMS: + tt = pos_params (varname, e1, e2, quoted); + if ((quoted & (Q_DOUBLE_QUOTES|Q_HERE_DOCUMENT)) == 0) + { + temp = tt ? quote_escapes (tt) : (char *)NULL; + FREE (tt); + } + else + temp = tt; + break; +#if defined (ARRAY_VARS) + case VT_ARRAYVAR: + if (assoc_p (v)) + /* we convert to list and take first e2 elements starting at e1th + element -- officially undefined for now */ + temp = assoc_subrange (assoc_cell (v), e1, e2, starsub, quoted); + else + /* We want E2 to be the number of elements desired (arrays can be sparse, + so verify_substring_values just returns the numbers specified and we + rely on array_subrange to understand how to deal with them). */ + temp = array_subrange (array_cell (v), e1, e2, starsub, quoted); + /* array_subrange now calls array_quote_escapes as appropriate, so the + caller no longer needs to. */ + break; +#endif + default: + temp = (char *)NULL; + } + + return temp; +} + +/****************************************************************/ +/* */ +/* Functions to perform pattern substitution on variable values */ +/* */ +/****************************************************************/ + +static int +shouldexp_replacement (s) + char *s; +{ + register char *p; + + for (p = s; p && *p; p++) + { + if (*p == '\\') + p++; + else if (*p == '&') + return 1; + } + return 0; +} + +char * +pat_subst (string, pat, rep, mflags) + char *string, *pat, *rep; + int mflags; +{ + char *ret, *s, *e, *str, *rstr, *mstr; + int rsize, rptr, l, replen, mtype, rxpand, rslen, mlen; + + if (string == 0) + return (savestring ("")); + + mtype = mflags & MATCH_TYPEMASK; + +#if 0 /* bash-4.2 ? */ + rxpand = (rep && *rep) ? shouldexp_replacement (rep) : 0; +#else + rxpand = 0; +#endif + + /* Special cases: + * 1. A null pattern with mtype == MATCH_BEG means to prefix STRING + * with REP and return the result. + * 2. A null pattern with mtype == MATCH_END means to append REP to + * STRING and return the result. + * These don't understand or process `&' in the replacement string. + */ + if ((pat == 0 || *pat == 0) && (mtype == MATCH_BEG || mtype == MATCH_END)) + { + replen = STRLEN (rep); + l = STRLEN (string); + ret = (char *)xmalloc (replen + l + 2); + if (replen == 0) + strcpy (ret, string); + else if (mtype == MATCH_BEG) + { + strcpy (ret, rep); + strcpy (ret + replen, string); + } + else + { + strcpy (ret, string); + strcpy (ret + l, rep); + } + return (ret); + } + + ret = (char *)xmalloc (rsize = 64); + ret[0] = '\0'; + + for (replen = STRLEN (rep), rptr = 0, str = string;;) + { + if (match_pattern (str, pat, mtype, &s, &e) == 0) + break; + l = s - str; + + if (rxpand) + { + int x; + mlen = e - s; + mstr = xmalloc (mlen + 1); + for (x = 0; x < mlen; x++) + mstr[x] = s[x]; + mstr[mlen] = '\0'; + rstr = strcreplace (rep, '&', mstr, 0); + rslen = strlen (rstr); + } + else + { + rstr = rep; + rslen = replen; + } + + RESIZE_MALLOCED_BUFFER (ret, rptr, (l + rslen), rsize, 64); + + /* OK, now copy the leading unmatched portion of the string (from + str to s) to ret starting at rptr (the current offset). Then copy + the replacement string at ret + rptr + (s - str). Increment + rptr (if necessary) and str and go on. */ + if (l) + { + strncpy (ret + rptr, str, l); + rptr += l; + } + if (replen) + { + strncpy (ret + rptr, rstr, rslen); + rptr += rslen; + } + str = e; /* e == end of match */ + + if (rstr != rep) + free (rstr); + + if (((mflags & MATCH_GLOBREP) == 0) || mtype != MATCH_ANY) + break; + + if (s == e) + { + /* On a zero-length match, make sure we copy one character, since + we increment one character to avoid infinite recursion. */ + RESIZE_MALLOCED_BUFFER (ret, rptr, 1, rsize, 64); + ret[rptr++] = *str++; + e++; /* avoid infinite recursion on zero-length match */ + } + } + + /* Now copy the unmatched portion of the input string */ + if (str && *str) + { + RESIZE_MALLOCED_BUFFER (ret, rptr, STRLEN(str) + 1, rsize, 64); + strcpy (ret + rptr, str); + } + else + ret[rptr] = '\0'; + + return ret; +} + +/* Do pattern match and replacement on the positional parameters. */ +static char * +pos_params_pat_subst (string, pat, rep, mflags) + char *string, *pat, *rep; + int mflags; +{ + WORD_LIST *save, *params; + WORD_DESC *w; + char *ret; + int pchar, qflags; + + save = params = list_rest_of_args (); + if (save == 0) + return ((char *)NULL); + + for ( ; params; params = params->next) + { + ret = pat_subst (params->word->word, pat, rep, mflags); + w = alloc_word_desc (); + w->word = ret ? ret : savestring (""); + dispose_word (params->word); + params->word = w; + } + + pchar = (mflags & MATCH_STARSUB) == MATCH_STARSUB ? '*' : '@'; + qflags = (mflags & MATCH_QUOTED) == MATCH_QUOTED ? Q_DOUBLE_QUOTES : 0; + + ret = string_list_pos_params (pchar, save, qflags); + + dispose_words (save); + + return (ret); +} + +/* Perform pattern substitution on VALUE, which is the expansion of + VARNAME. PATSUB is an expression supplying the pattern to match + and the string to substitute. QUOTED is a flags word containing + the type of quoting currently in effect. */ +static char * +parameter_brace_patsub (varname, value, ind, patsub, quoted, flags) + char *varname, *value; + int ind; + char *patsub; + int quoted, flags; +{ + int vtype, mflags, starsub, delim; + char *val, *temp, *pat, *rep, *p, *lpatsub, *tt; + SHELL_VAR *v; + + if (value == 0) + return ((char *)NULL); + + this_command_name = varname; + + vtype = get_var_and_type (varname, value, ind, quoted, flags, &v, &val); + if (vtype == -1) + return ((char *)NULL); + + starsub = vtype & VT_STARSUB; + vtype &= ~VT_STARSUB; + + mflags = 0; + /* PATSUB is never NULL when this is called. */ + if (*patsub == '/') + { + mflags |= MATCH_GLOBREP; + patsub++; + } + + /* Malloc this because expand_string_if_necessary or one of the expansion + functions in its call chain may free it on a substitution error. */ + lpatsub = savestring (patsub); + + if (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) + mflags |= MATCH_QUOTED; + + if (starsub) + mflags |= MATCH_STARSUB; + + /* If the pattern starts with a `/', make sure we skip over it when looking + for the replacement delimiter. */ + delim = skip_to_delim (lpatsub, ((*patsub == '/') ? 1 : 0), "/", 0); + if (lpatsub[delim] == '/') + { + lpatsub[delim] = 0; + rep = lpatsub + delim + 1; + } + else + rep = (char *)NULL; + + if (rep && *rep == '\0') + rep = (char *)NULL; + + /* Perform the same expansions on the pattern as performed by the + pattern removal expansions. */ + pat = getpattern (lpatsub, quoted, 1); + + if (rep) + { + /* We want to perform quote removal on the expanded replacement even if + the entire expansion is double-quoted because the parser and string + extraction functions treated quotes in the replacement string as + special. THIS IS NOT BACKWARDS COMPATIBLE WITH BASH-4.2. */ + if (shell_compatibility_level > 42) + rep = expand_string_if_necessary (rep, quoted & ~(Q_DOUBLE_QUOTES|Q_HERE_DOCUMENT), expand_string_unsplit); + /* This is the bash-4.2 code. */ + else if ((mflags & MATCH_QUOTED) == 0) + rep = expand_string_if_necessary (rep, quoted, expand_string_unsplit); + else + rep = expand_string_to_string_internal (rep, quoted, expand_string_unsplit); + } + + /* ksh93 doesn't allow the match specifier to be a part of the expanded + pattern. This is an extension. Make sure we don't anchor the pattern + at the beginning or end of the string if we're doing global replacement, + though. */ + p = pat; + if (mflags & MATCH_GLOBREP) + mflags |= MATCH_ANY; + else if (pat && pat[0] == '#') + { + mflags |= MATCH_BEG; + p++; + } + else if (pat && pat[0] == '%') + { + mflags |= MATCH_END; + p++; + } + else + mflags |= MATCH_ANY; + + /* OK, we now want to substitute REP for PAT in VAL. If + flags & MATCH_GLOBREP is non-zero, the substitution is done + everywhere, otherwise only the first occurrence of PAT is + replaced. The pattern matching code doesn't understand + CTLESC quoting CTLESC and CTLNUL so we use the dequoted variable + values passed in (VT_VARIABLE) so the pattern substitution + code works right. We need to requote special chars after + we're done for VT_VARIABLE and VT_ARRAYMEMBER, and for the + other cases if QUOTED == 0, since the posparams and arrays + indexed by * or @ do special things when QUOTED != 0. */ + + switch (vtype) + { + case VT_VARIABLE: + case VT_ARRAYMEMBER: + temp = pat_subst (val, p, rep, mflags); + if (vtype == VT_VARIABLE) + FREE (val); + if (temp) + { + tt = (mflags & MATCH_QUOTED) ? quote_string (temp) : quote_escapes (temp); + free (temp); + temp = tt; + } + break; + case VT_POSPARMS: + temp = pos_params_pat_subst (val, p, rep, mflags); + if (temp && (mflags & MATCH_QUOTED) == 0) + { + tt = quote_escapes (temp); + free (temp); + temp = tt; + } + break; +#if defined (ARRAY_VARS) + case VT_ARRAYVAR: + temp = assoc_p (v) ? assoc_patsub (assoc_cell (v), p, rep, mflags) + : array_patsub (array_cell (v), p, rep, mflags); + /* Don't call quote_escapes anymore; array_patsub calls + array_quote_escapes as appropriate before adding the + space separators; ditto for assoc_patsub. */ + break; +#endif + } + + FREE (pat); + FREE (rep); + free (lpatsub); + + return temp; +} + +/****************************************************************/ +/* */ +/* Functions to perform case modification on variable values */ +/* */ +/****************************************************************/ + +/* Do case modification on the positional parameters. */ + +static char * +pos_params_modcase (string, pat, modop, mflags) + char *string, *pat; + int modop; + int mflags; +{ + WORD_LIST *save, *params; + WORD_DESC *w; + char *ret; + int pchar, qflags; + + save = params = list_rest_of_args (); + if (save == 0) + return ((char *)NULL); + + for ( ; params; params = params->next) + { + ret = sh_modcase (params->word->word, pat, modop); + w = alloc_word_desc (); + w->word = ret ? ret : savestring (""); + dispose_word (params->word); + params->word = w; + } + + pchar = (mflags & MATCH_STARSUB) == MATCH_STARSUB ? '*' : '@'; + qflags = (mflags & MATCH_QUOTED) == MATCH_QUOTED ? Q_DOUBLE_QUOTES : 0; + + ret = string_list_pos_params (pchar, save, qflags); + dispose_words (save); + + return (ret); +} + +/* Perform case modification on VALUE, which is the expansion of + VARNAME. MODSPEC is an expression supplying the type of modification + to perform. QUOTED is a flags word containing the type of quoting + currently in effect. */ +static char * +parameter_brace_casemod (varname, value, ind, modspec, patspec, quoted, flags) + char *varname, *value; + int ind, modspec; + char *patspec; + int quoted, flags; +{ + int vtype, starsub, modop, mflags, x; + char *val, *temp, *pat, *p, *lpat, *tt; + SHELL_VAR *v; + + if (value == 0) + return ((char *)NULL); + + this_command_name = varname; + + vtype = get_var_and_type (varname, value, ind, quoted, flags, &v, &val); + if (vtype == -1) + return ((char *)NULL); + + starsub = vtype & VT_STARSUB; + vtype &= ~VT_STARSUB; + + modop = 0; + mflags = 0; + if (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) + mflags |= MATCH_QUOTED; + if (starsub) + mflags |= MATCH_STARSUB; + + p = patspec; + if (modspec == '^') + { + x = p && p[0] == modspec; + modop = x ? CASE_UPPER : CASE_UPFIRST; + p += x; + } + else if (modspec == ',') + { + x = p && p[0] == modspec; + modop = x ? CASE_LOWER : CASE_LOWFIRST; + p += x; + } + else if (modspec == '~') + { + x = p && p[0] == modspec; + modop = x ? CASE_TOGGLEALL : CASE_TOGGLE; + p += x; + } + + lpat = p ? savestring (p) : 0; + /* Perform the same expansions on the pattern as performed by the + pattern removal expansions. FOR LATER */ + pat = lpat ? getpattern (lpat, quoted, 1) : 0; + + /* OK, now we do the case modification. */ + switch (vtype) + { + case VT_VARIABLE: + case VT_ARRAYMEMBER: + temp = sh_modcase (val, pat, modop); + if (vtype == VT_VARIABLE) + FREE (val); + if (temp) + { + tt = (mflags & MATCH_QUOTED) ? quote_string (temp) : quote_escapes (temp); + free (temp); + temp = tt; + } + break; + + case VT_POSPARMS: + temp = pos_params_modcase (val, pat, modop, mflags); + if (temp && (mflags & MATCH_QUOTED) == 0) + { + tt = quote_escapes (temp); + free (temp); + temp = tt; + } + break; + +#if defined (ARRAY_VARS) + case VT_ARRAYVAR: + temp = assoc_p (v) ? assoc_modcase (assoc_cell (v), pat, modop, mflags) + : array_modcase (array_cell (v), pat, modop, mflags); + /* Don't call quote_escapes; array_modcase calls array_quote_escapes + as appropriate before adding the space separators; ditto for + assoc_modcase. */ + break; +#endif + } + + FREE (pat); + free (lpat); + + return temp; +} + +/* Check for unbalanced parens in S, which is the contents of $(( ... )). If + any occur, this must be a nested command substitution, so return 0. + Otherwise, return 1. A valid arithmetic expression must always have a + ( before a matching ), so any cases where there are more right parens + means that this must not be an arithmetic expression, though the parser + will not accept it without a balanced total number of parens. */ +static int +chk_arithsub (s, len) + const char *s; + int len; +{ + int i, count; + DECLARE_MBSTATE; + + i = count = 0; + while (i < len) + { + if (s[i] == LPAREN) + count++; + else if (s[i] == RPAREN) + { + count--; + if (count < 0) + return 0; + } + + switch (s[i]) + { + default: + ADVANCE_CHAR (s, len, i); + break; + + case '\\': + i++; + if (s[i]) + ADVANCE_CHAR (s, len, i); + break; + + case '\'': + i = skip_single_quoted (s, len, ++i); + break; + + case '"': + i = skip_double_quoted ((char *)s, len, ++i); + break; + } + } + + return (count == 0); +} + +/****************************************************************/ +/* */ +/* Functions to perform parameter expansion on a string */ +/* */ +/****************************************************************/ + +/* ${[#][!]name[[:][^[^]][,[,]]#[#]%[%]-=?+[word][:e1[:e2]]]} */ +static WORD_DESC * +parameter_brace_expand (string, indexp, quoted, pflags, quoted_dollar_atp, contains_dollar_at) + char *string; + int *indexp, quoted, *quoted_dollar_atp, *contains_dollar_at, pflags; +{ + int check_nullness, var_is_set, var_is_null, var_is_special; + int want_substring, want_indir, want_patsub, want_casemod; + char *name, *value, *temp, *temp1; + WORD_DESC *tdesc, *ret; + int t_index, sindex, c, tflag, modspec; + intmax_t number; + arrayind_t ind; + + temp = temp1 = value = (char *)NULL; + var_is_set = var_is_null = var_is_special = check_nullness = 0; + want_substring = want_indir = want_patsub = want_casemod = 0; + + sindex = *indexp; + t_index = ++sindex; + /* ${#var} doesn't have any of the other parameter expansions on it. */ + if (string[t_index] == '#' && legal_variable_starter (string[t_index+1])) /* {{ */ + name = string_extract (string, &t_index, "}", SX_VARNAME); + else +#if defined (CASEMOD_EXPANSIONS) + /* To enable case-toggling expansions using the `~' operator character + change the 1 to 0. */ +# if defined (CASEMOD_CAPCASE) + name = string_extract (string, &t_index, "#%^,~:-=?+/}", SX_VARNAME); +# else + name = string_extract (string, &t_index, "#%^,:-=?+/}", SX_VARNAME); +# endif /* CASEMOD_CAPCASE */ +#else + name = string_extract (string, &t_index, "#%:-=?+/}", SX_VARNAME); +#endif /* CASEMOD_EXPANSIONS */ + + ret = 0; + tflag = 0; + + ind = INTMAX_MIN; + + /* If the name really consists of a special variable, then make sure + that we have the entire name. We don't allow indirect references + to special variables except `#', `?', `@' and `*'. */ + if ((sindex == t_index && VALID_SPECIAL_LENGTH_PARAM (string[t_index])) || + (sindex == t_index - 1 && string[sindex] == '!' && VALID_INDIR_PARAM (string[t_index]))) + { + t_index++; + temp1 = string_extract (string, &t_index, "#%:-=?+/}", 0); + name = (char *)xrealloc (name, 3 + (strlen (temp1))); + *name = string[sindex]; + if (string[sindex] == '!') + { + /* indirect reference of $#, $?, $@, or $* */ + name[1] = string[sindex + 1]; + strcpy (name + 2, temp1); + } + else + strcpy (name + 1, temp1); + free (temp1); + } + sindex = t_index; + + /* Find out what character ended the variable name. Then + do the appropriate thing. */ + if (c = string[sindex]) + sindex++; + + /* If c is followed by one of the valid parameter expansion + characters, move past it as normal. If not, assume that + a substring specification is being given, and do not move + past it. */ + if (c == ':' && VALID_PARAM_EXPAND_CHAR (string[sindex])) + { + check_nullness++; + if (c = string[sindex]) + sindex++; + } + else if (c == ':' && string[sindex] != RBRACE) + want_substring = 1; + else if (c == '/' /* && string[sindex] != RBRACE */) /* XXX */ + want_patsub = 1; +#if defined (CASEMOD_EXPANSIONS) + else if (c == '^' || c == ',' || c == '~') + { + modspec = c; + want_casemod = 1; + } +#endif + + /* Catch the valid and invalid brace expressions that made it through the + tests above. */ + /* ${#-} is a valid expansion and means to take the length of $-. + Similarly for ${#?} and ${##}... */ + if (name[0] == '#' && name[1] == '\0' && check_nullness == 0 && + VALID_SPECIAL_LENGTH_PARAM (c) && string[sindex] == RBRACE) + { + name = (char *)xrealloc (name, 3); + name[1] = c; + name[2] = '\0'; + c = string[sindex++]; + } + + /* ...but ${#%}, ${#:}, ${#=}, ${#+}, and ${#/} are errors. */ + if (name[0] == '#' && name[1] == '\0' && check_nullness == 0 && + member (c, "%:=+/") && string[sindex] == RBRACE) + { + temp = (char *)NULL; + goto bad_substitution; + } + + /* Indirect expansion begins with a `!'. A valid indirect expansion is + either a variable name, one of the positional parameters or a special + variable that expands to one of the positional parameters. */ + want_indir = *name == '!' && + (legal_variable_starter ((unsigned char)name[1]) || DIGIT (name[1]) + || VALID_INDIR_PARAM (name[1])); + + /* Determine the value of this variable. */ + + /* Check for special variables, directly referenced. */ + if (SPECIAL_VAR (name, want_indir)) + var_is_special++; + + /* Check for special expansion things, like the length of a parameter */ + if (*name == '#' && name[1]) + { + /* If we are not pointing at the character just after the + closing brace, then we haven't gotten all of the name. + Since it begins with a special character, this is a bad + substitution. Also check NAME for validity before trying + to go on. */ + if (string[sindex - 1] != RBRACE || (valid_length_expression (name) == 0)) + { + temp = (char *)NULL; + goto bad_substitution; + } + + number = parameter_brace_expand_length (name); + if (number == INTMAX_MIN && unbound_vars_is_error) + { + last_command_exit_value = EXECUTION_FAILURE; + err_unboundvar (name+1); + free (name); + return (interactive_shell ? &expand_wdesc_error : &expand_wdesc_fatal); + } + free (name); + + *indexp = sindex; + if (number < 0) + return (&expand_wdesc_error); + else + { + ret = alloc_word_desc (); + ret->word = itos (number); + return ret; + } + } + + /* ${@} is identical to $@. */ + if (name[0] == '@' && name[1] == '\0') + { + if ((quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) && quoted_dollar_atp) + *quoted_dollar_atp = 1; + + if (contains_dollar_at) + *contains_dollar_at = 1; + + tflag |= W_DOLLARAT; + } + + /* Process ${!PREFIX*} expansion. */ + if (want_indir && string[sindex - 1] == RBRACE && + (string[sindex - 2] == '*' || string[sindex - 2] == '@') && + legal_variable_starter ((unsigned char) name[1])) + { + char **x; + WORD_LIST *xlist; + + temp1 = savestring (name + 1); + number = strlen (temp1); + temp1[number - 1] = '\0'; + x = all_variables_matching_prefix (temp1); + xlist = strvec_to_word_list (x, 0, 0); + if (string[sindex - 2] == '*') + temp = string_list_dollar_star (xlist); + else + { + temp = string_list_dollar_at (xlist, quoted); + if ((quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) && quoted_dollar_atp) + *quoted_dollar_atp = 1; + if (contains_dollar_at) + *contains_dollar_at = 1; + + tflag |= W_DOLLARAT; + } + free (x); + dispose_words (xlist); + free (temp1); + *indexp = sindex; + + free (name); + + ret = alloc_word_desc (); + ret->word = temp; + ret->flags = tflag; /* XXX */ + return ret; + } + +#if defined (ARRAY_VARS) + /* Process ${!ARRAY[@]} and ${!ARRAY[*]} expansion. */ /* [ */ + if (want_indir && string[sindex - 1] == RBRACE && + string[sindex - 2] == ']' && valid_array_reference (name+1)) + { + char *x, *x1; + + temp1 = savestring (name + 1); + x = array_variable_name (temp1, &x1, (int *)0); /* [ */ + FREE (x); + if (ALL_ELEMENT_SUB (x1[0]) && x1[1] == ']') + { + temp = array_keys (temp1, quoted); /* handles assoc vars too */ + if (x1[0] == '@') + { + if ((quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) && quoted_dollar_atp) + *quoted_dollar_atp = 1; + if (contains_dollar_at) + *contains_dollar_at = 1; + + tflag |= W_DOLLARAT; + } + + free (temp1); + *indexp = sindex; + + ret = alloc_word_desc (); + ret->word = temp; + ret->flags = tflag; /* XXX */ + return ret; + } + + free (temp1); + } +#endif /* ARRAY_VARS */ + + /* Make sure that NAME is valid before trying to go on. */ + if (valid_brace_expansion_word (want_indir ? name + 1 : name, + var_is_special) == 0) + { + temp = (char *)NULL; + goto bad_substitution; + } + + if (want_indir) + { + tdesc = parameter_brace_expand_indir (name + 1, var_is_special, quoted, quoted_dollar_atp, contains_dollar_at); + /* Turn off the W_ARRAYIND flag because there is no way for this function + to return the index we're supposed to be using. */ + if (tdesc && tdesc->flags) + tdesc->flags &= ~W_ARRAYIND; + } + else + tdesc = parameter_brace_expand_word (name, var_is_special, quoted, PF_IGNUNBOUND|(pflags&(PF_NOSPLIT2|PF_ASSIGNRHS)), &ind); + + if (tdesc) + { + temp = tdesc->word; + tflag = tdesc->flags; + dispose_word_desc (tdesc); + } + else + temp = (char *)0; + + if (temp == &expand_param_error || temp == &expand_param_fatal) + { + FREE (name); + FREE (value); + return (temp == &expand_param_error ? &expand_wdesc_error : &expand_wdesc_fatal); + } + +#if defined (ARRAY_VARS) + if (valid_array_reference (name)) + chk_atstar (name, quoted, quoted_dollar_atp, contains_dollar_at); +#endif + + var_is_set = temp != (char *)0; + var_is_null = check_nullness && (var_is_set == 0 || *temp == 0); + /* XXX - this may not need to be restricted to special variables */ + if (check_nullness) + var_is_null |= var_is_set && var_is_special && (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) && QUOTED_NULL (temp); + + /* Get the rest of the stuff inside the braces. */ + if (c && c != RBRACE) + { + /* Extract the contents of the ${ ... } expansion + according to the Posix.2 rules. */ + value = extract_dollar_brace_string (string, &sindex, quoted, (c == '%' || c == '#' || c =='/' || c == '^' || c == ',' || c ==':') ? SX_POSIXEXP|SX_WORD : SX_WORD); + if (string[sindex] == RBRACE) + sindex++; + else + goto bad_substitution; + } + else + value = (char *)NULL; + + *indexp = sindex; + + /* All the cases where an expansion can possibly generate an unbound + variable error. */ + if (want_substring || want_patsub || want_casemod || c == '#' || c == '%' || c == RBRACE) + { + if (var_is_set == 0 && unbound_vars_is_error && ((name[0] != '@' && name[0] != '*') || name[1])) + { + last_command_exit_value = EXECUTION_FAILURE; + err_unboundvar (name); + FREE (value); + FREE (temp); + free (name); + return (interactive_shell ? &expand_wdesc_error : &expand_wdesc_fatal); + } + } + + /* If this is a substring spec, process it and add the result. */ + if (want_substring) + { + temp1 = parameter_brace_substring (name, temp, ind, value, quoted, (tflag & W_ARRAYIND) ? AV_USEIND : 0); + FREE (name); + FREE (value); + FREE (temp); + + if (temp1 == &expand_param_error) + return (&expand_wdesc_error); + else if (temp1 == &expand_param_fatal) + return (&expand_wdesc_fatal); + + ret = alloc_word_desc (); + ret->word = temp1; + /* We test quoted_dollar_atp because we want variants with double-quoted + "$@" to take a different code path. In fact, we make sure at the end + of expand_word_internal that we're only looking at these flags if + quoted_dollar_at == 0. */ + if (temp1 && + (quoted_dollar_atp == 0 || *quoted_dollar_atp == 0) && + QUOTED_NULL (temp1) && (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES))) + ret->flags |= W_QUOTED|W_HASQUOTEDNULL; + return ret; + } + else if (want_patsub) + { + temp1 = parameter_brace_patsub (name, temp, ind, value, quoted, (tflag & W_ARRAYIND) ? AV_USEIND : 0); + FREE (name); + FREE (value); + FREE (temp); + + if (temp1 == &expand_param_error) + return (&expand_wdesc_error); + else if (temp1 == &expand_param_fatal) + return (&expand_wdesc_fatal); + + ret = alloc_word_desc (); + ret->word = temp1; + if (temp1 && + (quoted_dollar_atp == 0 || *quoted_dollar_atp == 0) && + QUOTED_NULL (temp1) && (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES))) + ret->flags |= W_QUOTED|W_HASQUOTEDNULL; + return ret; + } +#if defined (CASEMOD_EXPANSIONS) + else if (want_casemod) + { + temp1 = parameter_brace_casemod (name, temp, ind, modspec, value, quoted, (tflag & W_ARRAYIND) ? AV_USEIND : 0); + FREE (name); + FREE (value); + FREE (temp); + + if (temp1 == &expand_param_error) + return (&expand_wdesc_error); + else if (temp1 == &expand_param_fatal) + return (&expand_wdesc_fatal); + + ret = alloc_word_desc (); + ret->word = temp1; + if (temp1 && + (quoted_dollar_atp == 0 || *quoted_dollar_atp == 0) && + QUOTED_NULL (temp1) && (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES))) + ret->flags |= W_QUOTED|W_HASQUOTEDNULL; + return ret; + } +#endif + + /* Do the right thing based on which character ended the variable name. */ + switch (c) + { + default: + case '\0': + bad_substitution: + last_command_exit_value = EXECUTION_FAILURE; + report_error (_("%s: bad substitution"), string ? string : "??"); + FREE (value); + FREE (temp); + free (name); + return &expand_wdesc_error; + + case RBRACE: + break; + + case '#': /* ${param#[#]pattern} */ + case '%': /* ${param%[%]pattern} */ + if (value == 0 || *value == '\0' || temp == 0 || *temp == '\0') + { + FREE (value); + break; + } + temp1 = parameter_brace_remove_pattern (name, temp, ind, value, c, quoted, (tflag & W_ARRAYIND) ? AV_USEIND : 0); + free (temp); + free (value); + free (name); + + ret = alloc_word_desc (); + ret->word = temp1; + if (temp1 && QUOTED_NULL (temp1) && (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES))) + ret->flags |= W_QUOTED|W_HASQUOTEDNULL; + return ret; + + case '-': + case '=': + case '?': + case '+': + if (var_is_set && var_is_null == 0) + { + /* If the operator is `+', we don't want the value of the named + variable for anything, just the value of the right hand side. */ + if (c == '+') + { + /* XXX -- if we're double-quoted and the named variable is "$@", + we want to turn off any special handling of "$@" -- + we're not using it, so whatever is on the rhs applies. */ + if ((quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) && quoted_dollar_atp) + *quoted_dollar_atp = 0; + if (contains_dollar_at) + *contains_dollar_at = 0; + + FREE (temp); + if (value) + { + /* From Posix discussion on austin-group list. Issue 221 + requires that backslashes escaping `}' inside + double-quoted ${...} be removed. */ + if (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) + quoted |= Q_DOLBRACE; + ret = parameter_brace_expand_rhs (name, value, c, + quoted, + quoted_dollar_atp, + contains_dollar_at); + /* XXX - fix up later, esp. noting presence of + W_HASQUOTEDNULL in ret->flags */ + free (value); + } + else + temp = (char *)NULL; + } + else + { + FREE (value); + } + /* Otherwise do nothing; just use the value in TEMP. */ + } + else /* VAR not set or VAR is NULL. */ + { + FREE (temp); + temp = (char *)NULL; + if (c == '=' && var_is_special) + { + last_command_exit_value = EXECUTION_FAILURE; + report_error (_("$%s: cannot assign in this way"), name); + free (name); + free (value); + return &expand_wdesc_error; + } + else if (c == '?') + { + parameter_brace_expand_error (name, value); + return (interactive_shell ? &expand_wdesc_error : &expand_wdesc_fatal); + } + else if (c != '+') + { + /* XXX -- if we're double-quoted and the named variable is "$@", + we want to turn off any special handling of "$@" -- + we're not using it, so whatever is on the rhs applies. */ + if ((quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) && quoted_dollar_atp) + *quoted_dollar_atp = 0; + if (contains_dollar_at) + *contains_dollar_at = 0; + + /* From Posix discussion on austin-group list. Issue 221 requires + that backslashes escaping `}' inside double-quoted ${...} be + removed. */ + if (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) + quoted |= Q_DOLBRACE; + ret = parameter_brace_expand_rhs (name, value, c, quoted, + quoted_dollar_atp, + contains_dollar_at); + /* XXX - fix up later, esp. noting presence of + W_HASQUOTEDNULL in tdesc->flags */ + } + free (value); + } + + break; + } + free (name); + + if (ret == 0) + { + ret = alloc_word_desc (); + ret->flags = tflag; + ret->word = temp; + } + return (ret); +} + +/* Expand a single ${xxx} expansion. The braces are optional. When + the braces are used, parameter_brace_expand() does the work, + possibly calling param_expand recursively. */ +static WORD_DESC * +param_expand (string, sindex, quoted, expanded_something, + contains_dollar_at, quoted_dollar_at_p, had_quoted_null_p, + pflags) + char *string; + int *sindex, quoted, *expanded_something, *contains_dollar_at; + int *quoted_dollar_at_p, *had_quoted_null_p, pflags; +{ + char *temp, *temp1, uerror[3]; + int zindex, t_index, expok; + unsigned char c; + intmax_t number; + SHELL_VAR *var; + WORD_LIST *list; + WORD_DESC *tdesc, *ret; + int tflag; + + zindex = *sindex; + c = string[++zindex]; + + temp = (char *)NULL; + ret = tdesc = (WORD_DESC *)NULL; + tflag = 0; + + /* Do simple cases first. Switch on what follows '$'. */ + switch (c) + { + /* $0 .. $9? */ + case '0': + case '1': + case '2': + case '3': + case '4': + case '5': + case '6': + case '7': + case '8': + case '9': + temp1 = dollar_vars[TODIGIT (c)]; + if (unbound_vars_is_error && temp1 == (char *)NULL) + { + uerror[0] = '$'; + uerror[1] = c; + uerror[2] = '\0'; + last_command_exit_value = EXECUTION_FAILURE; + err_unboundvar (uerror); + return (interactive_shell ? &expand_wdesc_error : &expand_wdesc_fatal); + } + if (temp1) + temp = (*temp1 && (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES))) + ? quote_string (temp1) + : quote_escapes (temp1); + else + temp = (char *)NULL; + + break; + + /* $$ -- pid of the invoking shell. */ + case '$': + temp = itos (dollar_dollar_pid); + break; + + /* $# -- number of positional parameters. */ + case '#': + temp = itos (number_of_args ()); + break; + + /* $? -- return value of the last synchronous command. */ + case '?': + temp = itos (last_command_exit_value); + break; + + /* $- -- flags supplied to the shell on invocation or by `set'. */ + case '-': + temp = which_set_flags (); + break; + + /* $! -- Pid of the last asynchronous command. */ + case '!': + /* If no asynchronous pids have been created, expand to nothing. + If `set -u' has been executed, and no async processes have + been created, this is an expansion error. */ + if (last_asynchronous_pid == NO_PID) + { + if (expanded_something) + *expanded_something = 0; + temp = (char *)NULL; + if (unbound_vars_is_error) + { + uerror[0] = '$'; + uerror[1] = c; + uerror[2] = '\0'; + last_command_exit_value = EXECUTION_FAILURE; + err_unboundvar (uerror); + return (interactive_shell ? &expand_wdesc_error : &expand_wdesc_fatal); + } + } + else + temp = itos (last_asynchronous_pid); + break; + + /* The only difference between this and $@ is when the arg is quoted. */ + case '*': /* `$*' */ + list = list_rest_of_args (); + +#if 0 + /* According to austin-group posix proposal by Geoff Clare in + <20090505091501.GA10097@squonk.masqnet> of 5 May 2009: + + "The shell shall write a message to standard error and + immediately exit when it tries to expand an unset parameter + other than the '@' and '*' special parameters." + */ + + if (list == 0 && unbound_vars_is_error && (pflags & PF_IGNUNBOUND) == 0) + { + uerror[0] = '$'; + uerror[1] = '*'; + uerror[2] = '\0'; + last_command_exit_value = EXECUTION_FAILURE; + err_unboundvar (uerror); + return (interactive_shell ? &expand_wdesc_error : &expand_wdesc_fatal); + } +#endif + + /* If there are no command-line arguments, this should just + disappear if there are other characters in the expansion, + even if it's quoted. */ + if ((quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) && list == 0) + temp = (char *)NULL; + else if (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES|Q_PATQUOTE)) + { + /* If we have "$*" we want to make a string of the positional + parameters, separated by the first character of $IFS, and + quote the whole string, including the separators. If IFS + is unset, the parameters are separated by ' '; if $IFS is + null, the parameters are concatenated. */ + temp = (quoted & (Q_DOUBLE_QUOTES|Q_PATQUOTE)) ? string_list_dollar_star (list) : string_list (list); + if (temp) + { + temp1 = (quoted & Q_DOUBLE_QUOTES) ? quote_string (temp) : temp; + if (*temp == 0) + tflag |= W_HASQUOTEDNULL; + if (temp != temp1) + free (temp); + temp = temp1; + } + } + else + { + /* We check whether or not we're eventually going to split $* here, + for example when IFS is empty and we are processing the rhs of + an assignment statement. In that case, we don't separate the + arguments at all. Otherwise, if the $* is not quoted it is + identical to $@ */ +# if defined (HANDLE_MULTIBYTE) + if (expand_no_split_dollar_star && ifs_firstc[0] == 0) +# else + if (expand_no_split_dollar_star && ifs_firstc == 0) +# endif + temp = string_list_dollar_star (list); + else + { + temp = string_list_dollar_at (list, quoted); + if (quoted == 0 && (ifs_is_set == 0 || ifs_is_null)) + tflag |= W_SPLITSPACE; + } + + if (expand_no_split_dollar_star == 0 && contains_dollar_at) + *contains_dollar_at = 1; + } + + dispose_words (list); + break; + + /* When we have "$@" what we want is "$1" "$2" "$3" ... This + means that we have to turn quoting off after we split into + the individually quoted arguments so that the final split + on the first character of $IFS is still done. */ + case '@': /* `$@' */ + list = list_rest_of_args (); + +#if 0 + /* According to austin-group posix proposal by Geoff Clare in + <20090505091501.GA10097@squonk.masqnet> of 5 May 2009: + + "The shell shall write a message to standard error and + immediately exit when it tries to expand an unset parameter + other than the '@' and '*' special parameters." + */ + + if (list == 0 && unbound_vars_is_error && (pflags & PF_IGNUNBOUND) == 0) + { + uerror[0] = '$'; + uerror[1] = '@'; + uerror[2] = '\0'; + last_command_exit_value = EXECUTION_FAILURE; + err_unboundvar (uerror); + return (interactive_shell ? &expand_wdesc_error : &expand_wdesc_fatal); + } +#endif + + /* We want to flag the fact that we saw this. We can't turn + off quoting entirely, because other characters in the + string might need it (consider "\"$@\""), but we need some + way to signal that the final split on the first character + of $IFS should be done, even though QUOTED is 1. */ + /* XXX - should this test include Q_PATQUOTE? */ + if (quoted_dollar_at_p && (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES))) + *quoted_dollar_at_p = 1; + if (contains_dollar_at) + *contains_dollar_at = 1; + + /* We want to separate the positional parameters with the first + character of $IFS in case $IFS is something other than a space. + We also want to make sure that splitting is done no matter what -- + according to POSIX.2, this expands to a list of the positional + parameters no matter what IFS is set to. */ + /* XXX - what to do when in a context where word splitting is not + performed? Even when IFS is not the default, posix seems to imply + that we behave like unquoted $* ? Maybe we should use PF_NOSPLIT2 + here. */ + temp = string_list_dollar_at (list, (pflags & PF_ASSIGNRHS) ? (quoted|Q_DOUBLE_QUOTES) : quoted); + + tflag |= W_DOLLARAT; + dispose_words (list); + break; + + case LBRACE: + tdesc = parameter_brace_expand (string, &zindex, quoted, pflags, + quoted_dollar_at_p, + contains_dollar_at); + + if (tdesc == &expand_wdesc_error || tdesc == &expand_wdesc_fatal) + return (tdesc); + temp = tdesc ? tdesc->word : (char *)0; + + /* XXX */ + /* Quoted nulls should be removed if there is anything else + in the string. */ + /* Note that we saw the quoted null so we can add one back at + the end of this function if there are no other characters + in the string, discard TEMP, and go on. The exception to + this is when we have "${@}" and $1 is '', since $@ needs + special handling. */ + if (tdesc && tdesc->word && (tdesc->flags & W_HASQUOTEDNULL) && QUOTED_NULL (temp)) + { + if (had_quoted_null_p) + *had_quoted_null_p = 1; + if (*quoted_dollar_at_p == 0) + { + free (temp); + tdesc->word = temp = (char *)NULL; + } + + } + + ret = tdesc; + goto return0; + + /* Do command or arithmetic substitution. */ + case LPAREN: + /* We have to extract the contents of this paren substitution. */ + t_index = zindex + 1; + temp = extract_command_subst (string, &t_index, 0); + zindex = t_index; + + /* For Posix.2-style `$(( ))' arithmetic substitution, + extract the expression and pass it to the evaluator. */ + if (temp && *temp == LPAREN) + { + char *temp2; + temp1 = temp + 1; + temp2 = savestring (temp1); + t_index = strlen (temp2) - 1; + + if (temp2[t_index] != RPAREN) + { + free (temp2); + goto comsub; + } + + /* Cut off ending `)' */ + temp2[t_index] = '\0'; + + if (chk_arithsub (temp2, t_index) == 0) + { + free (temp2); +#if 0 + internal_warning (_("future versions of the shell will force evaluation as an arithmetic substitution")); +#endif + goto comsub; + } + + /* Expand variables found inside the expression. */ + temp1 = expand_arith_string (temp2, Q_DOUBLE_QUOTES|Q_ARITH); + free (temp2); + +arithsub: + /* No error messages. */ + this_command_name = (char *)NULL; + number = evalexp (temp1, &expok); + free (temp); + free (temp1); + if (expok == 0) + { + if (interactive_shell == 0 && posixly_correct) + { + last_command_exit_value = EXECUTION_FAILURE; + return (&expand_wdesc_fatal); + } + else + return (&expand_wdesc_error); + } + temp = itos (number); + break; + } + +comsub: + if (pflags & PF_NOCOMSUB) + /* we need zindex+1 because string[zindex] == RPAREN */ + temp1 = substring (string, *sindex, zindex+1); + else + { + tdesc = command_substitute (temp, quoted); + temp1 = tdesc ? tdesc->word : (char *)NULL; + if (tdesc) + dispose_word_desc (tdesc); + } + FREE (temp); + temp = temp1; + break; + + /* Do POSIX.2d9-style arithmetic substitution. This will probably go + away in a future bash release. */ + case '[': + /* Extract the contents of this arithmetic substitution. */ + t_index = zindex + 1; + temp = extract_arithmetic_subst (string, &t_index); + zindex = t_index; + if (temp == 0) + { + temp = savestring (string); + if (expanded_something) + *expanded_something = 0; + goto return0; + } + + /* Do initial variable expansion. */ + temp1 = expand_arith_string (temp, Q_DOUBLE_QUOTES|Q_ARITH); + + goto arithsub; + + default: + /* Find the variable in VARIABLE_LIST. */ + temp = (char *)NULL; + + for (t_index = zindex; (c = string[zindex]) && legal_variable_char (c); zindex++) + ; + temp1 = (zindex > t_index) ? substring (string, t_index, zindex) : (char *)NULL; + + /* If this isn't a variable name, then just output the `$'. */ + if (temp1 == 0 || *temp1 == '\0') + { + FREE (temp1); + temp = (char *)xmalloc (2); + temp[0] = '$'; + temp[1] = '\0'; + if (expanded_something) + *expanded_something = 0; + goto return0; + } + + /* If the variable exists, return its value cell. */ + var = find_variable (temp1); + + if (var && invisible_p (var) == 0 && var_isset (var)) + { +#if defined (ARRAY_VARS) + if (assoc_p (var) || array_p (var)) + { + temp = array_p (var) ? array_reference (array_cell (var), 0) + : assoc_reference (assoc_cell (var), "0"); + if (temp) + temp = (*temp && (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES))) + ? quote_string (temp) + : quote_escapes (temp); + else if (unbound_vars_is_error) + goto unbound_variable; + } + else +#endif + { + temp = value_cell (var); + + temp = (*temp && (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES))) + ? quote_string (temp) + : quote_escapes (temp); + } + + free (temp1); + + goto return0; + } + else if (var && (invisible_p (var) || var_isset (var) == 0)) + temp = (char *)NULL; + else if ((var = find_variable_last_nameref (temp1)) && var_isset (var) && invisible_p (var) == 0) + { + temp = nameref_cell (var); +#if defined (ARRAY_VARS) + if (temp && *temp && valid_array_reference (temp)) + { + tdesc = parameter_brace_expand_word (temp, SPECIAL_VAR (temp, 0), quoted, pflags, (arrayind_t *)NULL); + if (tdesc == &expand_wdesc_error || tdesc == &expand_wdesc_fatal) + return (tdesc); + ret = tdesc; + goto return0; + } + else +#endif + /* y=2 ; typeset -n x=y; echo $x is not the same as echo $2 in ksh */ + if (temp && *temp && legal_identifier (temp) == 0) + { + last_command_exit_value = EXECUTION_FAILURE; + report_error (_("%s: invalid variable name for name reference"), temp); + return (&expand_wdesc_error); /* XXX */ + } + else + temp = (char *)NULL; + } + + temp = (char *)NULL; + +unbound_variable: + if (unbound_vars_is_error) + { + last_command_exit_value = EXECUTION_FAILURE; + err_unboundvar (temp1); + } + else + { + free (temp1); + goto return0; + } + + free (temp1); + last_command_exit_value = EXECUTION_FAILURE; + return ((unbound_vars_is_error && interactive_shell == 0) + ? &expand_wdesc_fatal + : &expand_wdesc_error); + } + + if (string[zindex]) + zindex++; + +return0: + *sindex = zindex; + + if (ret == 0) + { + ret = alloc_word_desc (); + ret->flags = tflag; /* XXX */ + ret->word = temp; + } + return ret; +} + +/* Make a word list which is the result of parameter and variable + expansion, command substitution, arithmetic substitution, and + quote removal of WORD. Return a pointer to a WORD_LIST which is + the result of the expansion. If WORD contains a null word, the + word list returned is also null. + + QUOTED contains flag values defined in shell.h. + + ISEXP is used to tell expand_word_internal that the word should be + treated as the result of an expansion. This has implications for + how IFS characters in the word are treated. + + CONTAINS_DOLLAR_AT and EXPANDED_SOMETHING are return values; when non-null + they point to an integer value which receives information about expansion. + CONTAINS_DOLLAR_AT gets non-zero if WORD contained "$@", else zero. + EXPANDED_SOMETHING get non-zero if WORD contained any parameter expansions, + else zero. + + This only does word splitting in the case of $@ expansion. In that + case, we split on ' '. */ + +/* Values for the local variable quoted_state. */ +#define UNQUOTED 0 +#define PARTIALLY_QUOTED 1 +#define WHOLLY_QUOTED 2 + +static WORD_LIST * +expand_word_internal (word, quoted, isexp, contains_dollar_at, expanded_something) + WORD_DESC *word; + int quoted, isexp; + int *contains_dollar_at; + int *expanded_something; +{ + WORD_LIST *list; + WORD_DESC *tword; + + /* The intermediate string that we build while expanding. */ + char *istring; + + /* The current size of the above object. */ + int istring_size; + + /* Index into ISTRING. */ + int istring_index; + + /* Temporary string storage. */ + char *temp, *temp1; + + /* The text of WORD. */ + register char *string; + + /* The size of STRING. */ + size_t string_size; + + /* The index into STRING. */ + int sindex; + + /* This gets 1 if we see a $@ while quoted. */ + int quoted_dollar_at; + + /* One of UNQUOTED, PARTIALLY_QUOTED, or WHOLLY_QUOTED, depending on + whether WORD contains no quoting characters, a partially quoted + string (e.g., "xx"ab), or is fully quoted (e.g., "xxab"). */ + int quoted_state; + + /* State flags */ + int had_quoted_null; + int has_dollar_at, temp_has_dollar_at; + int split_on_spaces; + int tflag; + int pflags; /* flags passed to param_expand */ + + int assignoff; /* If assignment, offset of `=' */ + + register unsigned char c; /* Current character. */ + int t_index; /* For calls to string_extract_xxx. */ + + char twochars[2]; + + DECLARE_MBSTATE; + + istring = (char *)xmalloc (istring_size = DEFAULT_INITIAL_ARRAY_SIZE); + istring[istring_index = 0] = '\0'; + quoted_dollar_at = had_quoted_null = has_dollar_at = 0; + split_on_spaces = 0; + quoted_state = UNQUOTED; + + string = word->word; + if (string == 0) + goto finished_with_string; + /* Don't need the string length for the SADD... and COPY_ macros unless + multibyte characters are possible. */ + string_size = (MB_CUR_MAX > 1) ? strlen (string) : 1; + + if (contains_dollar_at) + *contains_dollar_at = 0; + + assignoff = -1; + + /* Begin the expansion. */ + + for (sindex = 0; ;) + { + c = string[sindex]; + + /* Case on top-level character. */ + switch (c) + { + case '\0': + goto finished_with_string; + + case CTLESC: + sindex++; +#if HANDLE_MULTIBYTE + if (MB_CUR_MAX > 1 && string[sindex]) + { + SADD_MBQCHAR_BODY(temp, string, sindex, string_size); + } + else +#endif + { + temp = (char *)xmalloc (3); + temp[0] = CTLESC; + temp[1] = c = string[sindex]; + temp[2] = '\0'; + } + +dollar_add_string: + if (string[sindex]) + sindex++; + +add_string: + if (temp) + { + istring = sub_append_string (temp, istring, &istring_index, &istring_size); + temp = (char *)0; + } + + break; + +#if defined (PROCESS_SUBSTITUTION) + /* Process substitution. */ + case '<': + case '>': + { + /* bash-4.4/bash-5.0 + XXX - technically this should only be expanded at the start + of a word */ + if (string[++sindex] != LPAREN || (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) || (word->flags & (W_DQUOTE|W_NOPROCSUB)) || posixly_correct) + { + sindex--; /* add_character: label increments sindex */ + goto add_character; + } + else + t_index = sindex + 1; /* skip past both '<' and LPAREN */ + + temp1 = extract_process_subst (string, (c == '<') ? "<(" : ">(", &t_index, 0); /*))*/ + sindex = t_index; + + /* If the process substitution specification is `<()', we want to + open the pipe for writing in the child and produce output; if + it is `>()', we want to open the pipe for reading in the child + and consume input. */ + temp = temp1 ? process_substitute (temp1, (c == '>')) : (char *)0; + + FREE (temp1); + + goto dollar_add_string; + } +#endif /* PROCESS_SUBSTITUTION */ + + case '=': + /* Posix.2 section 3.6.1 says that tildes following `=' in words + which are not assignment statements are not expanded. If the + shell isn't in posix mode, though, we perform tilde expansion + on `likely candidate' unquoted assignment statements (flags + include W_ASSIGNMENT but not W_QUOTED). A likely candidate + contains an unquoted :~ or =~. Something to think about: we + now have a flag that says to perform tilde expansion on arguments + to `assignment builtins' like declare and export that look like + assignment statements. We now do tilde expansion on such words + even in POSIX mode. */ + if (word->flags & (W_ASSIGNRHS|W_NOTILDE)) + { + if (isexp == 0 && (word->flags & (W_NOSPLIT|W_NOSPLIT2)) == 0 && isifs (c)) + goto add_ifs_character; + else + goto add_character; + } + /* If we're not in posix mode or forcing assignment-statement tilde + expansion, note where the `=' appears in the word and prepare to + do tilde expansion following the first `='. */ + if ((word->flags & W_ASSIGNMENT) && + (posixly_correct == 0 || (word->flags & W_TILDEEXP)) && + assignoff == -1 && sindex > 0) + assignoff = sindex; + if (sindex == assignoff && string[sindex+1] == '~') /* XXX */ + word->flags |= W_ITILDE; +#if 0 + else if ((word->flags & W_ASSIGNMENT) && + (posixly_correct == 0 || (word->flags & W_TILDEEXP)) && + string[sindex+1] == '~') + word->flags |= W_ITILDE; +#endif + if (isexp == 0 && (word->flags & (W_NOSPLIT|W_NOSPLIT2)) == 0 && isifs (c)) + goto add_ifs_character; + else + goto add_character; + + case ':': + if (word->flags & W_NOTILDE) + { + if (isexp == 0 && (word->flags & (W_NOSPLIT|W_NOSPLIT2)) == 0 && isifs (c)) + goto add_ifs_character; + else + goto add_character; + } + + if ((word->flags & (W_ASSIGNMENT|W_ASSIGNRHS|W_TILDEEXP)) && + string[sindex+1] == '~') + word->flags |= W_ITILDE; + + if (isexp == 0 && (word->flags & (W_NOSPLIT|W_NOSPLIT2)) == 0 && isifs (c)) + goto add_ifs_character; + else + goto add_character; + + case '~': + /* If the word isn't supposed to be tilde expanded, or we're not + at the start of a word or after an unquoted : or = in an + assignment statement, we don't do tilde expansion. If we don't want + tilde expansion when expanding words to be passed to the arithmetic + evaluator, remove the check for Q_ARITH. */ + if ((word->flags & (W_NOTILDE|W_DQUOTE)) || + (sindex > 0 && ((word->flags & W_ITILDE) == 0)) || + ((quoted & (Q_DOUBLE_QUOTES|Q_HERE_DOCUMENT)) && ((quoted & Q_ARITH) == 0))) + { + word->flags &= ~W_ITILDE; + if (isexp == 0 && (word->flags & (W_NOSPLIT|W_NOSPLIT2)) == 0 && isifs (c) && (quoted & (Q_DOUBLE_QUOTES|Q_HERE_DOCUMENT)) == 0) + goto add_ifs_character; + else + goto add_character; + } + + if (word->flags & W_ASSIGNRHS) + tflag = 2; + else if (word->flags & (W_ASSIGNMENT|W_TILDEEXP)) + tflag = 1; + else + tflag = 0; + + temp = bash_tilde_find_word (string + sindex, tflag, &t_index); + + word->flags &= ~W_ITILDE; + + if (temp && *temp && t_index > 0) + { + temp1 = bash_tilde_expand (temp, tflag); + if (temp1 && *temp1 == '~' && STREQ (temp, temp1)) + { + FREE (temp); + FREE (temp1); + goto add_character; /* tilde expansion failed */ + } + free (temp); + temp = temp1; + sindex += t_index; + goto add_quoted_string; /* XXX was add_string */ + } + else + { + FREE (temp); + goto add_character; + } + + case '$': + if (expanded_something) + *expanded_something = 1; + + temp_has_dollar_at = 0; + pflags = (word->flags & W_NOCOMSUB) ? PF_NOCOMSUB : 0; + if (word->flags & W_NOSPLIT2) + pflags |= PF_NOSPLIT2; + if (word->flags & W_ASSIGNRHS) + pflags |= PF_ASSIGNRHS; + tword = param_expand (string, &sindex, quoted, expanded_something, + &temp_has_dollar_at, "ed_dollar_at, + &had_quoted_null, pflags); + has_dollar_at += temp_has_dollar_at; + split_on_spaces += (tword->flags & W_SPLITSPACE); + + if (tword == &expand_wdesc_error || tword == &expand_wdesc_fatal) + { + free (string); + free (istring); + return ((tword == &expand_wdesc_error) ? &expand_word_error + : &expand_word_fatal); + } + if (contains_dollar_at && has_dollar_at) + *contains_dollar_at = 1; + + if (tword && (tword->flags & W_HASQUOTEDNULL)) + had_quoted_null = 1; + + temp = tword ? tword->word : (char *)NULL; + dispose_word_desc (tword); + + /* Kill quoted nulls; we will add them back at the end of + expand_word_internal if nothing else in the string */ + if (had_quoted_null && temp && QUOTED_NULL (temp)) + { + FREE (temp); + temp = (char *)NULL; + } + + goto add_string; + break; + + case '`': /* Backquoted command substitution. */ + { + t_index = sindex++; + + temp = string_extract (string, &sindex, "`", SX_REQMATCH); + /* The test of sindex against t_index is to allow bare instances of + ` to pass through, for backwards compatibility. */ + if (temp == &extract_string_error || temp == &extract_string_fatal) + { + if (sindex - 1 == t_index) + { + sindex = t_index; + goto add_character; + } + last_command_exit_value = EXECUTION_FAILURE; + report_error (_("bad substitution: no closing \"`\" in %s") , string+t_index); + free (string); + free (istring); + return ((temp == &extract_string_error) ? &expand_word_error + : &expand_word_fatal); + } + + if (expanded_something) + *expanded_something = 1; + + if (word->flags & W_NOCOMSUB) + /* sindex + 1 because string[sindex] == '`' */ + temp1 = substring (string, t_index, sindex + 1); + else + { + de_backslash (temp); + tword = command_substitute (temp, quoted); + temp1 = tword ? tword->word : (char *)NULL; + if (tword) + dispose_word_desc (tword); + } + FREE (temp); + temp = temp1; + goto dollar_add_string; + } + + case '\\': + if (string[sindex + 1] == '\n') + { + sindex += 2; + continue; + } + + c = string[++sindex]; + + if (quoted & Q_HERE_DOCUMENT) + tflag = CBSHDOC; + else if (quoted & Q_DOUBLE_QUOTES) + tflag = CBSDQUOTE; + else + tflag = 0; + + /* From Posix discussion on austin-group list: Backslash escaping + a } in ${...} is removed. Issue 0000221 */ + if ((quoted & Q_DOLBRACE) && c == RBRACE) + { + SCOPY_CHAR_I (twochars, CTLESC, c, string, sindex, string_size); + } + /* This is the fix for " $@\ " */ + else if ((quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) && ((sh_syntaxtab[c] & tflag) == 0) && isexp == 0 && isifs (c)) + { + RESIZE_MALLOCED_BUFFER (istring, istring_index, 2, istring_size, + DEFAULT_ARRAY_SIZE); + istring[istring_index++] = CTLESC; + istring[istring_index++] = '\\'; + istring[istring_index] = '\0'; + + SCOPY_CHAR_I (twochars, CTLESC, c, string, sindex, string_size); + } + else if ((quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) && ((sh_syntaxtab[c] & tflag) == 0)) + { + SCOPY_CHAR_I (twochars, '\\', c, string, sindex, string_size); + } + else if (c == 0) + { + c = CTLNUL; + sindex--; /* add_character: label increments sindex */ + goto add_character; + } + else + { + SCOPY_CHAR_I (twochars, CTLESC, c, string, sindex, string_size); + } + + sindex++; +add_twochars: + /* BEFORE jumping here, we need to increment sindex if appropriate */ + RESIZE_MALLOCED_BUFFER (istring, istring_index, 2, istring_size, + DEFAULT_ARRAY_SIZE); + istring[istring_index++] = twochars[0]; + istring[istring_index++] = twochars[1]; + istring[istring_index] = '\0'; + + break; + + case '"': + if ((quoted & (Q_DOUBLE_QUOTES|Q_HERE_DOCUMENT)) && ((quoted & Q_ARITH) == 0)) + goto add_character; + + t_index = ++sindex; + temp = string_extract_double_quoted (string, &sindex, 0); + + /* If the quotes surrounded the entire string, then the + whole word was quoted. */ + quoted_state = (t_index == 1 && string[sindex] == '\0') + ? WHOLLY_QUOTED + : PARTIALLY_QUOTED; + + if (temp && *temp) + { + tword = alloc_word_desc (); + tword->word = temp; + + temp = (char *)NULL; + + temp_has_dollar_at = 0; /* XXX */ + /* Need to get W_HASQUOTEDNULL flag through this function. */ + list = expand_word_internal (tword, Q_DOUBLE_QUOTES, 0, &temp_has_dollar_at, (int *)NULL); + has_dollar_at += temp_has_dollar_at; + + if (list == &expand_word_error || list == &expand_word_fatal) + { + free (istring); + free (string); + /* expand_word_internal has already freed temp_word->word + for us because of the way it prints error messages. */ + tword->word = (char *)NULL; + dispose_word (tword); + return list; + } + + dispose_word (tword); + + /* "$@" (a double-quoted dollar-at) expands into nothing, + not even a NULL word, when there are no positional + parameters. */ + if (list == 0 && has_dollar_at) + { + quoted_dollar_at++; + break; + } + + /* If we get "$@", we know we have expanded something, so we + need to remember it for the final split on $IFS. This is + a special case; it's the only case where a quoted string + can expand into more than one word. It's going to come back + from the above call to expand_word_internal as a list with + a single word, in which all characters are quoted and + separated by blanks. What we want to do is to turn it back + into a list for the next piece of code. */ + if (list) + dequote_list (list); + + if (list && list->word && (list->word->flags & W_HASQUOTEDNULL)) + had_quoted_null = 1; /* XXX */ + + if (has_dollar_at) + { + quoted_dollar_at++; + if (contains_dollar_at) + *contains_dollar_at = 1; + if (expanded_something) + *expanded_something = 1; + } + } + else + { + /* What we have is "". This is a minor optimization. */ + FREE (temp); + list = (WORD_LIST *)NULL; + } + + /* The code above *might* return a list (consider the case of "$@", + where it returns "$1", "$2", etc.). We can't throw away the + rest of the list, and we have to make sure each word gets added + as quoted. We test on tresult->next: if it is non-NULL, we + quote the whole list, save it to a string with string_list, and + add that string. We don't need to quote the results of this + (and it would be wrong, since that would quote the separators + as well), so we go directly to add_string. */ + if (list) + { + if (list->next) + { + /* Testing quoted_dollar_at makes sure that "$@" is + split correctly when $IFS does not contain a space. */ + temp = quoted_dollar_at + ? string_list_dollar_at (list, Q_DOUBLE_QUOTES) + : string_list (quote_list (list)); + dispose_words (list); + goto add_string; + } + else + { + temp = savestring (list->word->word); + tflag = list->word->flags; + dispose_words (list); + + /* If the string is not a quoted null string, we want + to remove any embedded unquoted CTLNUL characters. + We do not want to turn quoted null strings back into + the empty string, though. We do this because we + want to remove any quoted nulls from expansions that + contain other characters. For example, if we have + x"$*"y or "x$*y" and there are no positional parameters, + the $* should expand into nothing. */ + /* We use the W_HASQUOTEDNULL flag to differentiate the + cases: a quoted null character as above and when + CTLNUL is contained in the (non-null) expansion + of some variable. We use the had_quoted_null flag to + pass the value through this function to its caller. */ + if ((tflag & W_HASQUOTEDNULL) && QUOTED_NULL (temp) == 0) + remove_quoted_nulls (temp); /* XXX */ + } + } + else + temp = (char *)NULL; + + /* We do not want to add quoted nulls to strings that are only + partially quoted; we can throw them away. The exception to + this is when we are going to be performing word splitting, + since we have to preserve a null argument if the next character + will cause word splitting. */ + if (temp == 0 && quoted_state == PARTIALLY_QUOTED && (word->flags & (W_NOSPLIT|W_NOSPLIT2))) + continue; + + add_quoted_string: + + if (temp) + { + temp1 = temp; + temp = quote_string (temp); + free (temp1); + goto add_string; + } + else + { + /* Add NULL arg. */ + c = CTLNUL; + sindex--; /* add_character: label increments sindex */ + goto add_character; + } + + /* break; */ + + case '\'': + if ((quoted & (Q_DOUBLE_QUOTES|Q_HERE_DOCUMENT))) + goto add_character; + + t_index = ++sindex; + temp = string_extract_single_quoted (string, &sindex); + + /* If the entire STRING was surrounded by single quotes, + then the string is wholly quoted. */ + quoted_state = (t_index == 1 && string[sindex] == '\0') + ? WHOLLY_QUOTED + : PARTIALLY_QUOTED; + + /* If all we had was '', it is a null expansion. */ + if (*temp == '\0') + { + free (temp); + temp = (char *)NULL; + } + else + remove_quoted_escapes (temp); /* ??? */ + + /* We do not want to add quoted nulls to strings that are only + partially quoted; such nulls are discarded. */ + if (temp == 0 && (quoted_state == PARTIALLY_QUOTED)) + continue; + + /* If we have a quoted null expansion, add a quoted NULL to istring. */ + if (temp == 0) + { + c = CTLNUL; + sindex--; /* add_character: label increments sindex */ + goto add_character; + } + else + goto add_quoted_string; + + /* break; */ + + default: + /* This is the fix for " $@ " */ + add_ifs_character: + if ((quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) || (isexp == 0 && isifs (c))) + { + if (string[sindex]) /* from old goto dollar_add_string */ + sindex++; + if (c == 0) + { + c = CTLNUL; + goto add_character; + } + else + { +#if HANDLE_MULTIBYTE + if (MB_CUR_MAX > 1) + sindex--; + + if (MB_CUR_MAX > 1) + { + SADD_MBQCHAR_BODY(temp, string, sindex, string_size); + } + else +#endif + { + twochars[0] = CTLESC; + twochars[1] = c; + goto add_twochars; + } + } + } + + SADD_MBCHAR (temp, string, sindex, string_size); + + add_character: + RESIZE_MALLOCED_BUFFER (istring, istring_index, 1, istring_size, + DEFAULT_ARRAY_SIZE); + istring[istring_index++] = c; + istring[istring_index] = '\0'; + + /* Next character. */ + sindex++; + } + } + +finished_with_string: + /* OK, we're ready to return. If we have a quoted string, and + quoted_dollar_at is not set, we do no splitting at all; otherwise + we split on ' '. The routines that call this will handle what to + do if nothing has been expanded. */ + + /* Partially and wholly quoted strings which expand to the empty + string are retained as an empty arguments. Unquoted strings + which expand to the empty string are discarded. The single + exception is the case of expanding "$@" when there are no + positional parameters. In that case, we discard the expansion. */ + + /* Because of how the code that handles "" and '' in partially + quoted strings works, we need to make ISTRING into a QUOTED_NULL + if we saw quoting characters, but the expansion was empty. + "" and '' are tossed away before we get to this point when + processing partially quoted strings. This makes "" and $xxx"" + equivalent when xxx is unset. We also look to see whether we + saw a quoted null from a ${} expansion and add one back if we + need to. */ + + /* If we expand to nothing and there were no single or double quotes + in the word, we throw it away. Otherwise, we return a NULL word. + The single exception is for $@ surrounded by double quotes when + there are no positional parameters. In that case, we also throw + the word away. */ + + if (*istring == '\0') + { + if (quoted_dollar_at == 0 && (had_quoted_null || quoted_state == PARTIALLY_QUOTED)) + { + istring[0] = CTLNUL; + istring[1] = '\0'; + tword = make_bare_word (istring); + tword->flags |= W_HASQUOTEDNULL; /* XXX */ + list = make_word_list (tword, (WORD_LIST *)NULL); + if (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) + tword->flags |= W_QUOTED; + } + /* According to sh, ksh, and Posix.2, if a word expands into nothing + and a double-quoted "$@" appears anywhere in it, then the entire + word is removed. */ + else if (quoted_state == UNQUOTED || quoted_dollar_at) + list = (WORD_LIST *)NULL; +#if 0 + else + { + tword = make_bare_word (istring); + if (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) + tword->flags |= W_QUOTED; + list = make_word_list (tword, (WORD_LIST *)NULL); + } +#else + else + list = (WORD_LIST *)NULL; +#endif + } + else if (word->flags & W_NOSPLIT) + { + tword = make_bare_word (istring); + if (word->flags & W_ASSIGNMENT) + tword->flags |= W_ASSIGNMENT; /* XXX */ + if (word->flags & W_COMPASSIGN) + tword->flags |= W_COMPASSIGN; /* XXX */ + if (word->flags & W_NOGLOB) + tword->flags |= W_NOGLOB; /* XXX */ + if (word->flags & W_NOBRACE) + tword->flags |= W_NOBRACE; /* XXX */ + if (word->flags & W_NOEXPAND) + tword->flags |= W_NOEXPAND; /* XXX */ + if (quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) + tword->flags |= W_QUOTED; + if (had_quoted_null && QUOTED_NULL (istring)) + tword->flags |= W_HASQUOTEDNULL; + list = make_word_list (tword, (WORD_LIST *)NULL); + } + else + { + char *ifs_chars; + char *tstring; + + ifs_chars = (quoted_dollar_at || has_dollar_at) ? ifs_value : (char *)NULL; + + /* If we have $@, we need to split the results no matter what. If + IFS is unset or NULL, string_list_dollar_at has separated the + positional parameters with a space, so we split on space (we have + set ifs_chars to " \t\n" above if ifs is unset). If IFS is set, + string_list_dollar_at has separated the positional parameters + with the first character of $IFS, so we split on $IFS. If + SPLIT_ON_SPACES is set, we expanded $* (unquoted) with IFS either + unset or null, and we want to make sure that we split on spaces + regardless of what else has happened to IFS since the expansion. */ + if (split_on_spaces) + list = list_string (istring, " ", 1); /* XXX quoted == 1? */ + /* If we have $@ (has_dollar_at != 0) and we are in a context where we + don't want to split the result (W_NOSPLIT2), and we are not quoted, + we have already separated the arguments with the first character of + $IFS. In this case, we want to return a list with a single word + with the separator possibly replaced with a space (it's what other + shells seem to do). + quoted_dollar_at is internal to this function and is set if we are + passed an argument that is unquoted (quoted == 0) but we encounter a + double-quoted $@ while expanding it. */ + else if (has_dollar_at && quoted_dollar_at == 0 && ifs_chars && quoted == 0 && (word->flags & W_NOSPLIT2)) + { + /* Only split and rejoin if we have to */ + if (*ifs_chars && *ifs_chars != ' ') + { + list = list_string (istring, *ifs_chars ? ifs_chars : " ", 1); + tstring = string_list (list); + } + else + tstring = istring; + tword = make_bare_word (tstring); + if (tstring != istring) + free (tstring); + goto set_word_flags; + } + else if (has_dollar_at && ifs_chars) + list = list_string (istring, *ifs_chars ? ifs_chars : " ", 1); + else + { + tword = make_bare_word (istring); +set_word_flags: + if ((quoted & (Q_DOUBLE_QUOTES|Q_HERE_DOCUMENT)) || (quoted_state == WHOLLY_QUOTED)) + tword->flags |= W_QUOTED; + if (word->flags & W_ASSIGNMENT) + tword->flags |= W_ASSIGNMENT; + if (word->flags & W_COMPASSIGN) + tword->flags |= W_COMPASSIGN; + if (word->flags & W_NOGLOB) + tword->flags |= W_NOGLOB; + if (word->flags & W_NOBRACE) + tword->flags |= W_NOBRACE; + if (word->flags & W_NOEXPAND) + tword->flags |= W_NOEXPAND; + if (had_quoted_null && QUOTED_NULL (istring)) + tword->flags |= W_HASQUOTEDNULL; /* XXX */ + list = make_word_list (tword, (WORD_LIST *)NULL); + } + } + + free (istring); + return (list); +} + +/* **************************************************************** */ +/* */ +/* Functions for Quote Removal */ +/* */ +/* **************************************************************** */ + +/* Perform quote removal on STRING. If QUOTED > 0, assume we are obeying the + backslash quoting rules for within double quotes or a here document. */ +char * +string_quote_removal (string, quoted) + char *string; + int quoted; +{ + size_t slen; + char *r, *result_string, *temp, *send; + int sindex, tindex, dquote; + unsigned char c; + DECLARE_MBSTATE; + + /* The result can be no longer than the original string. */ + slen = strlen (string); + send = string + slen; + + r = result_string = (char *)xmalloc (slen + 1); + + for (dquote = sindex = 0; c = string[sindex];) + { + switch (c) + { + case '\\': + c = string[++sindex]; + if (c == 0) + { + *r++ = '\\'; + break; + } + if (((quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) || dquote) && (sh_syntaxtab[c] & CBSDQUOTE) == 0) + *r++ = '\\'; + /* FALLTHROUGH */ + + default: + SCOPY_CHAR_M (r, string, send, sindex); + break; + + case '\'': + if ((quoted & (Q_HERE_DOCUMENT|Q_DOUBLE_QUOTES)) || dquote) + { + *r++ = c; + sindex++; + break; + } + tindex = sindex + 1; + temp = string_extract_single_quoted (string, &tindex); + if (temp) + { + strcpy (r, temp); + r += strlen (r); + free (temp); + } + sindex = tindex; + break; + + case '"': + dquote = 1 - dquote; + sindex++; + break; + } + } + *r = '\0'; + return (result_string); +} + +#if 0 +/* UNUSED */ +/* Perform quote removal on word WORD. This allocates and returns a new + WORD_DESC *. */ +WORD_DESC * +word_quote_removal (word, quoted) + WORD_DESC *word; + int quoted; +{ + WORD_DESC *w; + char *t; + + t = string_quote_removal (word->word, quoted); + w = alloc_word_desc (); + w->word = t ? t : savestring (""); + return (w); +} + +/* Perform quote removal on all words in LIST. If QUOTED is non-zero, + the members of the list are treated as if they are surrounded by + double quotes. Return a new list, or NULL if LIST is NULL. */ +WORD_LIST * +word_list_quote_removal (list, quoted) + WORD_LIST *list; + int quoted; +{ + WORD_LIST *result, *t, *tresult, *e; + + for (t = list, result = (WORD_LIST *)NULL; t; t = t->next) + { + tresult = make_word_list (word_quote_removal (t->word, quoted), (WORD_LIST *)NULL); +#if 0 + result = (WORD_LIST *) list_append (result, tresult); +#else + if (result == 0) + result = e = tresult; + else + { + e->next = tresult; + while (e->next) + e = e->next; + } +#endif + } + return (result); +} +#endif + +/******************************************* + * * + * Functions to perform word splitting * + * * + *******************************************/ + +void +setifs (v) + SHELL_VAR *v; +{ + char *t; + unsigned char uc; + + ifs_var = v; + ifs_value = (v && value_cell (v)) ? value_cell (v) : " \t\n"; + + ifs_is_set = ifs_var != 0; + ifs_is_null = ifs_is_set && (*ifs_value == 0); + + /* Should really merge ifs_cmap with sh_syntaxtab. XXX - doesn't yet + handle multibyte chars in IFS */ + memset (ifs_cmap, '\0', sizeof (ifs_cmap)); + for (t = ifs_value ; t && *t; t++) + { + uc = *t; + ifs_cmap[uc] = 1; + } + +#if defined (HANDLE_MULTIBYTE) + if (ifs_value == 0) + { + ifs_firstc[0] = '\0'; + ifs_firstc_len = 1; + } + else + { + size_t ifs_len; + ifs_len = strnlen (ifs_value, MB_CUR_MAX); + ifs_firstc_len = MBLEN (ifs_value, ifs_len); + if (ifs_firstc_len == 1 || ifs_firstc_len == 0 || MB_INVALIDCH (ifs_firstc_len)) + { + ifs_firstc[0] = ifs_value[0]; + ifs_firstc[1] = '\0'; + ifs_firstc_len = 1; + } + else + memcpy (ifs_firstc, ifs_value, ifs_firstc_len); + } +#else + ifs_firstc = ifs_value ? *ifs_value : 0; +#endif +} + +char * +getifs () +{ + return ifs_value; +} + +/* This splits a single word into a WORD LIST on $IFS, but only if the word + is not quoted. list_string () performs quote removal for us, even if we + don't do any splitting. */ +WORD_LIST * +word_split (w, ifs_chars) + WORD_DESC *w; + char *ifs_chars; +{ + WORD_LIST *result; + + if (w) + { + char *xifs; + + xifs = ((w->flags & W_QUOTED) || ifs_chars == 0) ? "" : ifs_chars; + result = list_string (w->word, xifs, w->flags & W_QUOTED); + } + else + result = (WORD_LIST *)NULL; + + return (result); +} + +/* Perform word splitting on LIST and return the RESULT. It is possible + to return (WORD_LIST *)NULL. */ +static WORD_LIST * +word_list_split (list) + WORD_LIST *list; +{ + WORD_LIST *result, *t, *tresult, *e; + + for (t = list, result = (WORD_LIST *)NULL; t; t = t->next) + { + tresult = word_split (t->word, ifs_value); + if (result == 0) + result = e = tresult; + else + { + e->next = tresult; + while (e->next) + e = e->next; + } + } + return (result); +} + +/************************************************** + * * + * Functions to expand an entire WORD_LIST * + * * + **************************************************/ + +/* Do any word-expansion-specific cleanup and jump to top_level */ +static void +exp_jump_to_top_level (v) + int v; +{ + set_pipestatus_from_exit (last_command_exit_value); + + /* Cleanup code goes here. */ + expand_no_split_dollar_star = 0; /* XXX */ + expanding_redir = 0; + assigning_in_environment = 0; + + if (parse_and_execute_level == 0) + top_level_cleanup (); /* from sig.c */ + + jump_to_top_level (v); +} + +/* Put NLIST (which is a WORD_LIST * of only one element) at the front of + ELIST, and set ELIST to the new list. */ +#define PREPEND_LIST(nlist, elist) \ + do { nlist->next = elist; elist = nlist; } while (0) + +/* Separate out any initial variable assignments from TLIST. If set -k has + been executed, remove all assignment statements from TLIST. Initial + variable assignments and other environment assignments are placed + on SUBST_ASSIGN_VARLIST. */ +static WORD_LIST * +separate_out_assignments (tlist) + WORD_LIST *tlist; +{ + register WORD_LIST *vp, *lp; + + if (tlist == 0) + return ((WORD_LIST *)NULL); + + if (subst_assign_varlist) + dispose_words (subst_assign_varlist); /* Clean up after previous error */ + + subst_assign_varlist = (WORD_LIST *)NULL; + vp = lp = tlist; + + /* Separate out variable assignments at the start of the command. + Loop invariant: vp->next == lp + Loop postcondition: + lp = list of words left after assignment statements skipped + tlist = original list of words + */ + while (lp && (lp->word->flags & W_ASSIGNMENT)) + { + vp = lp; + lp = lp->next; + } + + /* If lp != tlist, we have some initial assignment statements. + We make SUBST_ASSIGN_VARLIST point to the list of assignment + words and TLIST point to the remaining words. */ + if (lp != tlist) + { + subst_assign_varlist = tlist; + /* ASSERT(vp->next == lp); */ + vp->next = (WORD_LIST *)NULL; /* terminate variable list */ + tlist = lp; /* remainder of word list */ + } + + /* vp == end of variable list */ + /* tlist == remainder of original word list without variable assignments */ + if (!tlist) + /* All the words in tlist were assignment statements */ + return ((WORD_LIST *)NULL); + + /* ASSERT(tlist != NULL); */ + /* ASSERT((tlist->word->flags & W_ASSIGNMENT) == 0); */ + + /* If the -k option is in effect, we need to go through the remaining + words, separate out the assignment words, and place them on + SUBST_ASSIGN_VARLIST. */ + if (place_keywords_in_env) + { + WORD_LIST *tp; /* tp == running pointer into tlist */ + + tp = tlist; + lp = tlist->next; + + /* Loop Invariant: tp->next == lp */ + /* Loop postcondition: tlist == word list without assignment statements */ + while (lp) + { + if (lp->word->flags & W_ASSIGNMENT) + { + /* Found an assignment statement, add this word to end of + subst_assign_varlist (vp). */ + if (!subst_assign_varlist) + subst_assign_varlist = vp = lp; + else + { + vp->next = lp; + vp = lp; + } + + /* Remove the word pointed to by LP from TLIST. */ + tp->next = lp->next; + /* ASSERT(vp == lp); */ + lp->next = (WORD_LIST *)NULL; + lp = tp->next; + } + else + { + tp = lp; + lp = lp->next; + } + } + } + return (tlist); +} + +#define WEXP_VARASSIGN 0x001 +#define WEXP_BRACEEXP 0x002 +#define WEXP_TILDEEXP 0x004 +#define WEXP_PARAMEXP 0x008 +#define WEXP_PATHEXP 0x010 + +/* All of the expansions, including variable assignments at the start of + the list. */ +#define WEXP_ALL (WEXP_VARASSIGN|WEXP_BRACEEXP|WEXP_TILDEEXP|WEXP_PARAMEXP|WEXP_PATHEXP) + +/* All of the expansions except variable assignments at the start of + the list. */ +#define WEXP_NOVARS (WEXP_BRACEEXP|WEXP_TILDEEXP|WEXP_PARAMEXP|WEXP_PATHEXP) + +/* All of the `shell expansions': brace expansion, tilde expansion, parameter + expansion, command substitution, arithmetic expansion, word splitting, and + quote removal. */ +#define WEXP_SHELLEXP (WEXP_BRACEEXP|WEXP_TILDEEXP|WEXP_PARAMEXP) + +/* Take the list of words in LIST and do the various substitutions. Return + a new list of words which is the expanded list, and without things like + variable assignments. */ + +WORD_LIST * +expand_words (list) + WORD_LIST *list; +{ + return (expand_word_list_internal (list, WEXP_ALL)); +} + +/* Same as expand_words (), but doesn't hack variable or environment + variables. */ +WORD_LIST * +expand_words_no_vars (list) + WORD_LIST *list; +{ + return (expand_word_list_internal (list, WEXP_NOVARS)); +} + +WORD_LIST * +expand_words_shellexp (list) + WORD_LIST *list; +{ + return (expand_word_list_internal (list, WEXP_SHELLEXP)); +} + +static WORD_LIST * +glob_expand_word_list (tlist, eflags) + WORD_LIST *tlist; + int eflags; +{ + char **glob_array, *temp_string; + register int glob_index; + WORD_LIST *glob_list, *output_list, *disposables, *next; + WORD_DESC *tword; + + output_list = disposables = (WORD_LIST *)NULL; + glob_array = (char **)NULL; + while (tlist) + { + /* For each word, either globbing is attempted or the word is + added to orig_list. If globbing succeeds, the results are + added to orig_list and the word (tlist) is added to the list + of disposable words. If globbing fails and failed glob + expansions are left unchanged (the shell default), the + original word is added to orig_list. If globbing fails and + failed glob expansions are removed, the original word is + added to the list of disposable words. orig_list ends up + in reverse order and requires a call to REVERSE_LIST to + be set right. After all words are examined, the disposable + words are freed. */ + next = tlist->next; + + /* If the word isn't an assignment and contains an unquoted + pattern matching character, then glob it. */ + if ((tlist->word->flags & W_NOGLOB) == 0 && + unquoted_glob_pattern_p (tlist->word->word)) + { + glob_array = shell_glob_filename (tlist->word->word); + + /* Handle error cases. + I don't think we should report errors like "No such file + or directory". However, I would like to report errors + like "Read failed". */ + + if (glob_array == 0 || GLOB_FAILED (glob_array)) + { + glob_array = (char **)xmalloc (sizeof (char *)); + glob_array[0] = (char *)NULL; + } + + /* Dequote the current word in case we have to use it. */ + if (glob_array[0] == NULL) + { + temp_string = dequote_string (tlist->word->word); + free (tlist->word->word); + tlist->word->word = temp_string; + } + + /* Make the array into a word list. */ + glob_list = (WORD_LIST *)NULL; + for (glob_index = 0; glob_array[glob_index]; glob_index++) + { + tword = make_bare_word (glob_array[glob_index]); + glob_list = make_word_list (tword, glob_list); + } + + if (glob_list) + { + output_list = (WORD_LIST *)list_append (glob_list, output_list); + PREPEND_LIST (tlist, disposables); + } + else if (fail_glob_expansion != 0) + { + last_command_exit_value = EXECUTION_FAILURE; + report_error (_("no match: %s"), tlist->word->word); + exp_jump_to_top_level (DISCARD); + } + else if (allow_null_glob_expansion == 0) + { + /* Failed glob expressions are left unchanged. */ + PREPEND_LIST (tlist, output_list); + } + else + { + /* Failed glob expressions are removed. */ + PREPEND_LIST (tlist, disposables); + } + } + else + { + /* Dequote the string. */ + temp_string = dequote_string (tlist->word->word); + free (tlist->word->word); + tlist->word->word = temp_string; + PREPEND_LIST (tlist, output_list); + } + + strvec_dispose (glob_array); + glob_array = (char **)NULL; + + tlist = next; + } + + if (disposables) + dispose_words (disposables); + + if (output_list) + output_list = REVERSE_LIST (output_list, WORD_LIST *); + + return (output_list); +} + +#if defined (BRACE_EXPANSION) +static WORD_LIST * +brace_expand_word_list (tlist, eflags) + WORD_LIST *tlist; + int eflags; +{ + register char **expansions; + char *temp_string; + WORD_LIST *disposables, *output_list, *next; + WORD_DESC *w; + int eindex; + + for (disposables = output_list = (WORD_LIST *)NULL; tlist; tlist = next) + { + next = tlist->next; + + if (tlist->word->flags & W_NOBRACE) + { +/*itrace("brace_expand_word_list: %s: W_NOBRACE", tlist->word->word);*/ + PREPEND_LIST (tlist, output_list); + continue; + } + + if ((tlist->word->flags & (W_COMPASSIGN|W_ASSIGNARG)) == (W_COMPASSIGN|W_ASSIGNARG)) + { +/*itrace("brace_expand_word_list: %s: W_COMPASSIGN|W_ASSIGNARG", tlist->word->word);*/ + PREPEND_LIST (tlist, output_list); + continue; + } + + /* Only do brace expansion if the word has a brace character. If + not, just add the word list element to BRACES and continue. In + the common case, at least when running shell scripts, this will + degenerate to a bunch of calls to `mbschr', and then what is + basically a reversal of TLIST into BRACES, which is corrected + by a call to REVERSE_LIST () on BRACES when the end of TLIST + is reached. */ + if (mbschr (tlist->word->word, LBRACE)) + { + expansions = brace_expand (tlist->word->word); + + for (eindex = 0; temp_string = expansions[eindex]; eindex++) + { + w = alloc_word_desc (); + w->word = temp_string; + + /* If brace expansion didn't change the word, preserve + the flags. We may want to preserve the flags + unconditionally someday -- XXX */ + if (STREQ (temp_string, tlist->word->word)) + w->flags = tlist->word->flags; + else + w = make_word_flags (w, temp_string); + + output_list = make_word_list (w, output_list); + } + free (expansions); + + /* Add TLIST to the list of words to be freed after brace + expansion has been performed. */ + PREPEND_LIST (tlist, disposables); + } + else + PREPEND_LIST (tlist, output_list); + } + + if (disposables) + dispose_words (disposables); + + if (output_list) + output_list = REVERSE_LIST (output_list, WORD_LIST *); + + return (output_list); +} +#endif + +#if defined (ARRAY_VARS) +/* Take WORD, a compound associative array assignment, and internally run + 'declare -A w', where W is the variable name portion of WORD. */ +static int +make_internal_declare (word, option) + char *word; + char *option; +{ + int t; + WORD_LIST *wl; + WORD_DESC *w; + + w = make_word (word); + + t = assignment (w->word, 0); + w->word[t] = '\0'; + + wl = make_word_list (w, (WORD_LIST *)NULL); + wl = make_word_list (make_word (option), wl); + + return (declare_builtin (wl)); +} +#endif + +static WORD_LIST * +shell_expand_word_list (tlist, eflags) + WORD_LIST *tlist; + int eflags; +{ + WORD_LIST *expanded, *orig_list, *new_list, *next, *temp_list; + int expanded_something, has_dollar_at; + char *temp_string; + + /* We do tilde expansion all the time. This is what 1003.2 says. */ + new_list = (WORD_LIST *)NULL; + for (orig_list = tlist; tlist; tlist = next) + { + temp_string = tlist->word->word; + + next = tlist->next; + +#if defined (ARRAY_VARS) + /* If this is a compound array assignment to a builtin that accepts + such assignments (e.g., `declare'), take the assignment and perform + it separately, handling the semantics of declarations inside shell + functions. This avoids the double-evaluation of such arguments, + because `declare' does some evaluation of compound assignments on + its own. */ + if ((tlist->word->flags & (W_COMPASSIGN|W_ASSIGNARG)) == (W_COMPASSIGN|W_ASSIGNARG)) + { + int t; + char opts[8], opti; + + opti = 0; + if (tlist->word->flags & (W_ASSIGNASSOC|W_ASSNGLOBAL|W_ASSIGNARRAY)) + opts[opti++] = '-'; + + if ((tlist->word->flags & (W_ASSIGNASSOC|W_ASSNGLOBAL)) == (W_ASSIGNASSOC|W_ASSNGLOBAL)) + { + opts[opti++] = 'g'; + opts[opti++] = 'A'; + } + else if (tlist->word->flags & W_ASSIGNASSOC) + opts[opti++] = 'A'; + else if ((tlist->word->flags & (W_ASSIGNARRAY|W_ASSNGLOBAL)) == (W_ASSIGNARRAY|W_ASSNGLOBAL)) + { + opts[opti++] = 'g'; + opts[opti++] = 'a'; + } + else if (tlist->word->flags & W_ASSIGNARRAY) + opts[opti++] = 'a'; + else if (tlist->word->flags & W_ASSNGLOBAL) + opts[opti++] = 'g'; + +#if 0 + /* If we have special handling note the integer attribute */ + if (opti > 0 && (tlist->word->flags & W_ASSIGNINT)) + opts[opti++] = 'i'; +#endif + + opts[opti] = '\0'; + if (opti > 0) + make_internal_declare (tlist->word->word, opts); + + t = do_word_assignment (tlist->word, 0); + if (t == 0) + { + last_command_exit_value = EXECUTION_FAILURE; + exp_jump_to_top_level (DISCARD); + } + + /* Now transform the word as ksh93 appears to do and go on */ + t = assignment (tlist->word->word, 0); + tlist->word->word[t] = '\0'; + tlist->word->flags &= ~(W_ASSIGNMENT|W_NOSPLIT|W_COMPASSIGN|W_ASSIGNARG|W_ASSIGNASSOC|W_ASSIGNARRAY); + } +#endif + + expanded_something = 0; + expanded = expand_word_internal + (tlist->word, 0, 0, &has_dollar_at, &expanded_something); + + if (expanded == &expand_word_error || expanded == &expand_word_fatal) + { + /* By convention, each time this error is returned, + tlist->word->word has already been freed. */ + tlist->word->word = (char *)NULL; + + /* Dispose our copy of the original list. */ + dispose_words (orig_list); + /* Dispose the new list we're building. */ + dispose_words (new_list); + + last_command_exit_value = EXECUTION_FAILURE; + if (expanded == &expand_word_error) + exp_jump_to_top_level (DISCARD); + else + exp_jump_to_top_level (FORCE_EOF); + } + + /* Don't split words marked W_NOSPLIT. */ + if (expanded_something && (tlist->word->flags & W_NOSPLIT) == 0) + { + temp_list = word_list_split (expanded); + dispose_words (expanded); + } + else + { + /* If no parameter expansion, command substitution, process + substitution, or arithmetic substitution took place, then + do not do word splitting. We still have to remove quoted + null characters from the result. */ + word_list_remove_quoted_nulls (expanded); + temp_list = expanded; + } + + expanded = REVERSE_LIST (temp_list, WORD_LIST *); + new_list = (WORD_LIST *)list_append (expanded, new_list); + } + + if (orig_list) + dispose_words (orig_list); + + if (new_list) + new_list = REVERSE_LIST (new_list, WORD_LIST *); + + return (new_list); +} + +/* The workhorse for expand_words () and expand_words_no_vars (). + First arg is LIST, a WORD_LIST of words. + Second arg EFLAGS is a flags word controlling which expansions are + performed. + + This does all of the substitutions: brace expansion, tilde expansion, + parameter expansion, command substitution, arithmetic expansion, + process substitution, word splitting, and pathname expansion, according + to the bits set in EFLAGS. Words with the W_QUOTED or W_NOSPLIT bits + set, or for which no expansion is done, do not undergo word splitting. + Words with the W_NOGLOB bit set do not undergo pathname expansion; words + with W_NOBRACE set do not undergo brace expansion (see + brace_expand_word_list above). */ +static WORD_LIST * +expand_word_list_internal (list, eflags) + WORD_LIST *list; + int eflags; +{ + WORD_LIST *new_list, *temp_list; + int tint; + + tempenv_assign_error = 0; + if (list == 0) + return ((WORD_LIST *)NULL); + + garglist = new_list = copy_word_list (list); + if (eflags & WEXP_VARASSIGN) + { + garglist = new_list = separate_out_assignments (new_list); + if (new_list == 0) + { + if (subst_assign_varlist) + { + /* All the words were variable assignments, so they are placed + into the shell's environment. */ + for (temp_list = subst_assign_varlist; temp_list; temp_list = temp_list->next) + { + this_command_name = (char *)NULL; /* no arithmetic errors */ + tint = do_word_assignment (temp_list->word, 0); + /* Variable assignment errors in non-interactive shells + running in Posix.2 mode cause the shell to exit. */ + if (tint == 0) + { + last_command_exit_value = EXECUTION_FAILURE; + if (interactive_shell == 0 && posixly_correct) + exp_jump_to_top_level (FORCE_EOF); + else + exp_jump_to_top_level (DISCARD); + } + } + dispose_words (subst_assign_varlist); + subst_assign_varlist = (WORD_LIST *)NULL; + } + return ((WORD_LIST *)NULL); + } + } + + /* Begin expanding the words that remain. The expansions take place on + things that aren't really variable assignments. */ + +#if defined (BRACE_EXPANSION) + /* Do brace expansion on this word if there are any brace characters + in the string. */ + if ((eflags & WEXP_BRACEEXP) && brace_expansion && new_list) + new_list = brace_expand_word_list (new_list, eflags); +#endif /* BRACE_EXPANSION */ + + /* Perform the `normal' shell expansions: tilde expansion, parameter and + variable substitution, command substitution, arithmetic expansion, + and word splitting. */ + new_list = shell_expand_word_list (new_list, eflags); + + /* Okay, we're almost done. Now let's just do some filename + globbing. */ + if (new_list) + { + if ((eflags & WEXP_PATHEXP) && disallow_filename_globbing == 0) + /* Glob expand the word list unless globbing has been disabled. */ + new_list = glob_expand_word_list (new_list, eflags); + else + /* Dequote the words, because we're not performing globbing. */ + new_list = dequote_list (new_list); + } + + if ((eflags & WEXP_VARASSIGN) && subst_assign_varlist) + { + sh_wassign_func_t *assign_func; + int is_special_builtin, is_builtin_or_func; + + /* If the remainder of the words expand to nothing, Posix.2 requires + that the variable and environment assignments affect the shell's + environment. */ + assign_func = new_list ? assign_in_env : do_word_assignment; + tempenv_assign_error = 0; + + is_builtin_or_func = (new_list && new_list->word && (find_shell_builtin (new_list->word->word) || find_function (new_list->word->word))); + /* Posix says that special builtins exit if a variable assignment error + occurs in an assignment preceding it. */ + is_special_builtin = (posixly_correct && new_list && new_list->word && find_special_builtin (new_list->word->word)); + + for (temp_list = subst_assign_varlist; temp_list; temp_list = temp_list->next) + { + this_command_name = (char *)NULL; + assigning_in_environment = (assign_func == assign_in_env); + tint = (*assign_func) (temp_list->word, is_builtin_or_func); + assigning_in_environment = 0; + /* Variable assignment errors in non-interactive shells running + in Posix.2 mode cause the shell to exit. */ + if (tint == 0) + { + if (assign_func == do_word_assignment) + { + last_command_exit_value = EXECUTION_FAILURE; + if (interactive_shell == 0 && posixly_correct && is_special_builtin) + exp_jump_to_top_level (FORCE_EOF); + else + exp_jump_to_top_level (DISCARD); + } + else + tempenv_assign_error++; + } + } + + dispose_words (subst_assign_varlist); + subst_assign_varlist = (WORD_LIST *)NULL; + } + + return (new_list); +} diff --git a/tests/RUN-ONE-TEST b/tests/RUN-ONE-TEST index 3efcf32d6..72ec06a2c 100755 --- a/tests/RUN-ONE-TEST +++ b/tests/RUN-ONE-TEST @@ -1,4 +1,4 @@ -BUILD_DIR=/usr/local/build/chet/bash/bash-current +BUILD_DIR=/usr/local/build/bash/bash-current THIS_SH=$BUILD_DIR/bash PATH=$PATH:$BUILD_DIR diff --git a/tests/RUN-ONE-TEST~ b/tests/RUN-ONE-TEST~ new file mode 100755 index 000000000..3efcf32d6 --- /dev/null +++ b/tests/RUN-ONE-TEST~ @@ -0,0 +1,9 @@ +BUILD_DIR=/usr/local/build/chet/bash/bash-current +THIS_SH=$BUILD_DIR/bash +PATH=$PATH:$BUILD_DIR + +export THIS_SH PATH + +rm -f /tmp/xx + +/bin/sh "$@" diff --git a/tests/dollar-at-star b/tests/dollar-at-star index 39724fbf0..1ae92e58f 100755 --- a/tests/dollar-at-star +++ b/tests/dollar-at-star @@ -214,8 +214,8 @@ ${THIS_SH} ./dollar-at-star1.sub # more tests for expanding $@ and $* in a context where there is no word # splitting ${THIS_SH} ./dollar-at-star2.sub - ${THIS_SH} ./dollar-at-star3.sub +${THIS_SH} ./dollar-at-star4.sub # tests for special expansion of "$*" and "${array[*]}" when used with other # expansions -- bugs through bash-2.05b diff --git a/tests/dollar-at-star4.sub b/tests/dollar-at-star4.sub new file mode 100644 index 000000000..59472b5bb --- /dev/null +++ b/tests/dollar-at-star4.sub @@ -0,0 +1,84 @@ +# tests for quoted and unquoted, split and unsplit expansions of $@ and $* +# Posix interpretation 221 speaks to this issue + +set -- a b +IFS= +x=abc + +printf "<%s>\n" ${x#$*} + +a="a bc" +echo ${a#$*} +unset x a + +a=$@ +echo "$a" + +a=$* +echo "$a" + +printf '<%s>' $@ ; echo + +IFS=' +' + +a=$@ +echo "$a" + +a=$* +echo "$a" + +unset a + +unset IFS +set a b "c d" +printf '<%s>' $* ; echo + +IFS= +set a b "c d" +printf '<%s>' $* ; echo + +IFS=: +set a b +printf '<%s>' ${foo=$*} ; echo +printf '<%s>' "${foo}" ; echo + +unset foo +IFS=' +' +printf '<%s>' ${foo=$@} ; echo +printf '<%s>' "$foo" ; echo + +IFS=? +set a c +foo=abcd + +echo ${foo#"$*"} +echo "${foo#$*}" + +echo "${foo#'a?c'}" +echo "${foo#a?c}" + +IFS=? +set a c +str=a$'\001'c +pat=a$'\001' +echo "${str#$pat}" + +set a b +IFS=: + +a=$@ +recho "$a" + +recho ${foo=$*} +recho "$foo" + +unset foo +IFS=' +' +recho ${foo=$@} +recho "$foo" + +shift $# +unset foo diff --git a/tests/dollar-at-star4.sub~ b/tests/dollar-at-star4.sub~ new file mode 100644 index 000000000..c0145371d --- /dev/null +++ b/tests/dollar-at-star4.sub~ @@ -0,0 +1,81 @@ +# tests for quoted and unquoted, split and unsplit expansions of $@ and $* +# Posix interpretation 221 speaks to this issue + +set -- a b +IFS= +x=abc + +printf "<%s>\n" ${x#$*} + +a="a bc" +echo ${a#$*} +unset x a + +a=$@ +echo "$a" + +a=$* +echo "$a" + +printf '<%s>' $@ ; echo + +IFS=' +' + +a=$@ +echo "$a" + +a=$* +echo "$a" + +unset a + +unset IFS +set a b "c d" +printf '<%s>' $* ; echo + +IFS= +set a b "c d" +printf '<%s>' $* ; echo + +IFS=: +set a b +printf '<%s>' ${foo=$*} ; echo +printf '<%s>' "${foo}" ; echo + +unset foo +IFS=' +' +printf '<%s>' ${foo=$@} ; echo +printf '<%s>' "$foo" ; echo + +IFS=? +set a c +foo=abcd + +echo ${foo#"$*"} +echo "${foo#$*}" + +echo "${foo#'a?c'}" +echo "${foo#a?c}" + +IFS=? +set a c +str=a$'\001'c +pat=a$'\001' +echo "${str#$pat}" + +set a b +IFS=: + +a=$@ +recho "$a" + +recho ${foo=$*} +recho "$foo" + +unset foo +IFS=' +' +recho ${foo=$@} +recho "$foo" diff --git a/tests/dollar-at-star~ b/tests/dollar-at-star~ new file mode 100755 index 000000000..39724fbf0 --- /dev/null +++ b/tests/dollar-at-star~ @@ -0,0 +1,266 @@ +# first, let's start with the basics + +recho "$@" +recho "$*" + +recho $@ +recho $* + +set a b + +recho "$*" + +# If IFS is null, the parameters are joined without separators +IFS='' +recho "$*" + +# If IFS is unset, the parameters are separated by spaces +unset IFS +recho "${*}" + +recho "$@" +recho $@ + +IFS='/' +set bob 'tom dick harry' joe +set $* +recho $# +recho $1 +recho $2 +recho $3 + +set bob 'tom dick harry' joe +set ${*} +recho $# +recho $1 +recho $2 +recho $3 + +set bob 'tom dick harry' joe +set $@ +recho $# +recho $1 +recho $2 +recho $3 + +set bob 'tom dick harry' joe +set ${@} +recho $# +recho $1 +recho $2 +recho $3 + +# according to POSIX.2, unquoted $* should expand to multiple words if +# $IFS is null, just like unquoted $@ +IFS='' +set bob 'tom dick harry' joe +set $* +recho $# +recho $1 +recho $2 +recho $3 + +set bob 'tom dick harry' joe +set $@ +recho $# +recho $1 +recho $2 +recho $3 + +# if IFS is unset, the individual positional parameters are split on +# " \t\n" if $* or $@ are unquoted +unset IFS +set bob 'tom dick harry' joe +set $* +recho $# +recho $1 +recho $2 +recho $3 + +set bob 'tom dick harry' joe +set $@ +recho $# +recho $1 +recho $2 +recho $3 + +# but not for "$@" or "$*" +set bob 'tom dick harry' joe +set "$*" +recho $# +recho $1 +recho $2 +recho $3 + +set bob 'tom dick harry' joe +set "$@" +recho $# +recho $1 +recho $2 +recho $3 + +# POSIX.2 says these should both expand the positional parameters +# to multiple words +set a b c d e +IFS="" +recho $@ +recho "$@" + +# this example is straight from the POSIX.2 rationale +set foo bar bam + +recho "$@" +recho "$*" + +unset IFS + +recho "$@" +recho $@ +recho "$*" + +IFS=: + +# special variables +set -- 1 2 3 4 5 6 7 8 9 10 + +bar=${*} +foo=$* +echo foo = "$foo" +echo bar = "$bar" + +foo1=$@ +bar1=${@} + +echo foo1 = "$foo1" +echo bar1 = "$bar1" + +foo2="$*" +bar2="${*}" + +echo foo2 = "$foo2" +echo bar2 = "$bar2" + +eval foo3='$*' bar3='${*}' +echo foo3 = "$foo3" +echo bar3 = "$bar3" + +case $* in +*\:*) echo ok 1;; +*) echo bad 1;; +esac + +case $@ in +*\:*) echo bad 2;; +*) echo ok 2;; +esac + +case "$*" in +*\:*) echo ok 3;; +*) echo bad 3;; +esac + +case "$@" in +*\:*) echo bad 4;; +*) echo ok 4;; +esac + +IFS=$' \t\n' + +bar=${*} +foo=$* +echo foo = "$foo" +echo bar = "$bar" + +foo1=$@ +bar1=${@} + +echo foo1 = "$foo1" +echo bar1 = "$bar1" + +foo2="$*" +bar2="${*}" + +echo foo2 = "$foo2" +echo bar2 = "$bar2" + +eval foo3='$*' bar3='${*}' +echo foo3 = "$foo3" +echo bar3 = "$bar3" + +case $* in +*\ *) echo ok 1;; +*) echo bad 1;; +esac + +case $@ in +*\ *) echo ok 2;; +*) echo bad 2;; +esac + +case "$*" in +*\ *) echo ok 3;; +*) echo bad 3;; +esac + +case "$@" in +*\ *) echo ok 4;; +*) echo bad 4;; +esac + +# tests for the effect of quoting $* and $@ in an assignment context (plus +# arrays) -- bugs through bash 4.2 +${THIS_SH} ./dollar-at-star1.sub + +# more tests for expanding $@ and $* in a context where there is no word +# splitting +${THIS_SH} ./dollar-at-star2.sub + +${THIS_SH} ./dollar-at-star3.sub + +# tests for special expansion of "$*" and "${array[*]}" when used with other +# expansions -- bugs through bash-2.05b +${THIS_SH} ./dollar-star1.sub + +# tests for expansion of "$@" on rhs of things like ${param:+word}. Bugs +# though bash-2.05b +${THIS_SH} ./dollar-at1.sub + +# tests for expansion of other variables in double-quoted strings containing +# $@. Bugs through bash-2.05b +${THIS_SH} ./dollar-at2.sub + +# tests for various expansions of $* in different contexts -- word split, +# no splitting, etc. when $IFS is NUL +${THIS_SH} ./dollar-star2.sub + +# tests for expansions of "${array[*]}" and "${array[@]}" when $IFS is not the +# default and the array contains null elements +${THIS_SH} ./dollar-star3.sub + +# test for set -u and expansions of $@ when there are no positional parameters +${THIS_SH} ./dollar-at3.sub +# test for set -u and expansions of $* when there are no positional parameters +${THIS_SH} ./dollar-star4.sub + +# tests for expansions of $* when IFS is null +${THIS_SH} ./dollar-star5.sub + +# tests for inappropriate word splitting through bash-4.2 +${THIS_SH} ./dollar-at4.sub + +# tests for problems with "$@" preceded and followed by other quoted expansions +# through bash-4.2 +${THIS_SH} ./dollar-at5.sub + +# tests for problems with "${@:1}" and other expansions with null entries +# in positional parameters +${THIS_SH} ./dollar-at6.sub + +# tests for expansions of $* when $1 == ""; problem through bash-4.2 +${THIS_SH} ./dollar-star6.sub + +# tests for expansions of $* (unquoted) when IFS changes (e.g., ${IFS:=-}) +# problem through bash-4.2 +${THIS_SH} ./dollar-star7.sub + +exit 0 diff --git a/tests/dollar.right b/tests/dollar.right index ea50477de..4918da43c 100644 --- a/tests/dollar.right +++ b/tests/dollar.right @@ -161,6 +161,30 @@ argv[1] = argv[1] = <123 abc> <123 abc> + +a bc +a b +ab + +a b +a b + + + + + + +abcd +d +abcd +d +c +argv[1] = +argv[1] = +argv[1] = +argv[1] = +argv[2] = +argv[1] = xa|xb|xc xa|xb|xc a|b|c diff --git a/tests/misc/regress/log.orig b/tests/misc/regress/log.orig new file mode 100644 index 000000000..c1f1e1991 --- /dev/null +++ b/tests/misc/regress/log.orig @@ -0,0 +1,50 @@ +:; ./shx + +sh: +<&$fd ok +nlbq Mon Aug 3 02:45:00 EDT 1992 +bang geoff +quote 712824302 +setbq defmsgid=<1992Aug3.024502.6176@host> +bgwait sleep done... wait 6187 + + +bash: +<&$fd ok +nlbq Mon Aug 3 02:45:09 EDT 1992 +bang geoff +quote 712824311 +setbq defmsgid=<1992Aug3.024512.6212@host> +bgwait sleep done... wait 6223 + + +ash: +<&$fd shx1: 4: Syntax error: Bad fd number +nlbq Mon Aug 3 02:45:19 EDT 1992 +bang geoff +quote getdate: `"now"' not a valid date + +setbq defmsgid=<1992Aug3.` echo 024521 +bgwait sleep done... wait 6241 + + +ksh: +<&$fd ok +nlbq ./shx: 6248 Memory fault - core dumped +bang geoff +quote getdate: `"now"' not a valid date + +setbq defmsgid=<1992Aug3.024530.6257@host> +bgwait no such job: 6265 +wait 6265 +sleep done... + +zsh: +<&$fd ok +nlbq Mon Aug 3 02:45:36 EDT 1992 +bang shx3: event not found: /s/ [4] +quote 712824337 +setbq defmsgid=<..6290@host> +bgwait shx7: unmatched " [9] +sleep done... +:; diff --git a/tests/misc/regress/shx.orig b/tests/misc/regress/shx.orig new file mode 100644 index 000000000..4b3bf2b82 --- /dev/null +++ b/tests/misc/regress/shx.orig @@ -0,0 +1,10 @@ +#! /bin/sh +for cmd in sh bash ash ksh zsh +do + echo + echo $cmd: + for demo in shx? + do + $cmd $demo + done +done